Ping An Technology (Shenzhen) Co., LTD.

Chine

Retour au propriétaire

1-100 de 4 933 pour Ping An Technology (Shenzhen) Co., LTD. Trier par
Recheche Texte
Brevet
International - WIPO
Affiner par Reset Report
Date
Nouveautés (dernières 4 semaines) 24
2022 janvier (MACJ) 14
2021 décembre 20
2021 novembre 73
2021 octobre 69
Voir plus
Classe IPC
G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales 386
G06F 17/30 - Recherche documentaire; Structures de bases de données à cet effet 372
G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques 308
H04L 29/06 - Commande de la communication; Traitement de la communication caractérisés par un protocole 227
H04L 29/08 - Procédure de commande de la transmission, p.ex. procédure de commande du niveau de la liaison 213
Voir plus
Résultats pour  brevets
  1     2     3     ...     50        Prochaine page

1.

QUESTION AND ANSWER LIBRARY CONSTRUCTION METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2021096633
Numéro de publication 2022/016995
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2022-01-27
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Li, Xiaojuan
  • Xu, Guoqiang

Abrégé

The present application relates to the technical field of data processing, and provides a question and answer library construction method and apparatus, and an electronic device and a storage medium. The method comprises: acquiring customer service conversation statements; according to a pre-set keyword, performing statement extraction on the customer service conversation statements to obtain a question statement; performing vector conversion on the question statement to obtain a statement vector; according to a pre-set particle vector and the statement vector, performing incremental clustering on the question statement to determine a category corresponding to the question statement; and according to the question statement and the category corresponding to the question statement, constructing a customer service question and answer library. According to the present application, the speed of constructing a question and answer library can be improved. In addition, the present application relates to the technical field of data processing and artificial intelligence. The present application also relates to blockchain technology, and the content of a question and answer library can be stored in a blockchain node. In addition, the present application can also be used in the field of smart government affairs, smart medical treatment or smart education, thus promoting the construction of smart cities.

Classes IPC  ?

2.

METHOD FOR PROCESSING INTERNET OF THINGS CONTROL SYSTEM, AND SYSTEM, SERVER, TERMINAL AND MEDIUM

      
Numéro d'application CN2021097111
Numéro de publication 2022/016998
Statut Délivré - en vigueur
Date de dépôt 2021-05-31
Date de publication 2022-01-27
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Xie, Wei

Abrégé

The present disclosure relates to the technical field of the Internet of Things and is mainly applied to Internet of Things scenarios such as smart logistics in a smart city. Specifically, disclosed are a method for processing an Internet of Things control system, a cloud server of an Internet of Things control system, a terminal, and a computer-readable storage medium. The method comprises: a system configuration module acquiring a system global parameter, wherein the system global parameter comprises a drive starting sequence and hardware connection information of virtual drives, and a node starting sequence of virtual nodes; the system configuration module controlling a corresponding virtual drive to start according to the drive starting sequence, and being used for controlling a corresponding virtual node to start according to the node starting sequence; and the virtual drive, after starting, being used for determining a corresponding physical hardware device according to the hardware connection information, and establishing a communication connection with the corresponding physical hardware device.

Classes IPC  ?

  • H04L 29/08 - Procédure de commande de la transmission, p.ex. procédure de commande du niveau de la liaison
  • G06F 9/4401 - Amorçage
  • H04L 29/06 - Commande de la communication; Traitement de la communication caractérisés par un protocole

3.

IMAGE PROCESSING METHOD, DEVICE, ELECTRONIC APPARATUS, AND COMPUTER READABLE STORAGE MEDIUM

      
Numéro d'application CN2021096713
Numéro de publication 2022/016996
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2022-01-27
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zheng, Ziqi
  • Xu, Guoqiang
  • Qiu, Han

Abrégé

Embodiments of the present application provide an image processing method, a device, an electronic apparatus, and a computer readable storage medium. The method comprises: acquiring a face image set comprising a face image of a first subject, a face image of a second subject in a specified pose, and a face image of a third subject having a specified facial expression; performing feature extraction on the face images in the face image set to obtain a first face feature set, the first face feature set comprising a face feature of the first subject, a pose feature of the second subject, and a facial expression feature of the third subject; and performing face synthesis according to the first face feature set to obtain a synthesized face image of the first subject, wherein the synthesized face image of the first subject has the face feature of the first subject, the pose feature of the second subject, and the facial expression feature of the third subject. The present application improves the quality of generated face images. In addition, the present application further relates to a blockchain technique, and the synthesized face image of the first subject can be written into a blockchain.

Classes IPC  ?

  • G06T 11/00 - Génération d'images bidimensionnelles [2D]
  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales

4.

AI RECOGNITION-BASED MEETING MINUTES GENERATION METHOD AND APPARATUS, DEVICE AND MEDIUM

      
Numéro d'application CN2021096433
Numéro de publication 2022/016994
Statut Délivré - en vigueur
Date de dépôt 2021-05-27
Date de publication 2022-01-27
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liu, Lu
  • Peng, Tao

Abrégé

Provided are an artificial intelligence (AI) recognition-based meeting minutes generation method and apparatus, an electronic device and a computer-readable storage medium, which relate to the technical field of AI. The method comprises: when a meeting minutes generation request is received, acquiring a list of participants from the meeting minutes generation request (S10); from within a configuration database, extracting voiceprint features corresponding to each participant in the list of participants (S11); when an audio detection instruction is received, determining an audio device that issues the audio detection instruction, and acquiring an audio to be recognized from within the audio device (S12); in order to prevent a plurality of participants from speaking at the same time and causing voice overlap, processing the audio to obtain audio features (S13); comparing the audio features to the voiceprint features to obtain comparison results, and performing comprehensive analysis according to the comparison results and binding results of the audio device to accurately determine a speaking subject (S14); performing voice recognition on the audio to obtain text information (S15); translating the text information into target language text corresponding to a preset language type (S16); and generating meeting minutes according to the speaking subject, the text information and the target language text without needing to additionally provide translators (S17). The meeting minutes may also be stored on a blockchain.

Classes IPC  ?

  • G10L 15/22 - Procédures utilisées pendant le processus de reconnaissance de la parole, p.ex. dialogue homme-machine 
  • G10L 17/02 - Opérations de prétraitement, p.ex. sélection de segment; Représentation ou modélisation de motifs, p.ex. fondée sur l’analyse linéaire discriminante [LDA] ou les composantes principales; Sélection ou extraction des caractéristiques
  • H04L 12/18 - Dispositions pour la fourniture de services particuliers aux abonnés pour la diffusion ou les conférences
  • H04N 7/15 - Systèmes pour conférences

5.

TRANSACTION DATA PROCESSING METHOD AND APPARATUS, AND COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020135606
Numéro de publication 2022/011947
Statut Délivré - en vigueur
Date de dépôt 2020-12-11
Date de publication 2022-01-20
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Ren, Jie

Abrégé

A transaction data processing method and an apparatus, and a computer device and a storage medium, the method comprising: on the basis of a plurality of pieces of transaction data, determining transaction objects contained therein, in order to determine main transaction objects and additional transaction objects and construct a directed graph; on the basis of the Tarjan algorithm, searching for strongly connected components in the directed graph in order to obtain a plurality of additional transaction object sets, and sequentially grouping the plurality of additional transaction object sets with each main transaction object respectively, to obtain a plurality of transaction object groups; and determining whether or not each transaction object group satisfies pre-set conditions, and when a group satisfies the conditions, then on the basis of the pre-set conditions, executing a data operation corresponding to the pre-set conditions, the pre-set conditions comprising a risk control condition and/or an information push condition. In addition, the present method further relates to blockchain technology, and the transaction data can be stored in a blockchain. The present method is able to implement transaction risk control and transaction object group recommendation.

Classes IPC  ?

  • G06Q 20/40 - Autorisation, p.ex. identification du payeur ou du bénéficiaire, vérification des références du client ou du magasin; Examen et approbation des payeurs, p.ex. contrôle des lignes de crédit ou des listes négatives

6.

MEDICAL TREATMENT ALLOCATION METHOD AND APPARATUS BASED ON RISK PREDICTION, AND COMPUTER DEVICE

      
Numéro d'application CN2021097209
Numéro de publication 2022/012188
Statut Délivré - en vigueur
Date de dépôt 2021-05-31
Date de publication 2022-01-20
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Tang, Rui

Abrégé

A medical treatment allocation method and apparatus based on risk prediction, and a computer device and a storage medium. The method comprises: according to prediction request information from a client, sending corresponding query information to the client, and acquiring reply information fed back; quantifying illness information therein according to an information quantification rule, so as to obtain corresponding illness quantification information; quantifying association information therein to obtain corresponding association information; acquiring a corresponding illness risk value according to the illness quantification information, acquiring a corresponding association risk value according to an association risk, and then performing integration to obtain a risk prediction result; and acquiring corresponding medical treatment allocation information according to the risk prediction result and a prediction type, and feeding the corresponding medical treatment allocation information back to the client. The method is based on intelligent decision technology, and belongs to the technical field of artificial intelligence. The illness risk of a customer can be evaluated by means of two aspects of information, thereby improving the accuracy of an obtained risk prediction result, and thus improving the efficiency and accuracy of performing medical treatment allocation on a patient.

Classes IPC  ?

  • G16H 40/20 - TIC spécialement adaptées à la gestion ou à l’administration de ressources ou d’établissements de santé; TIC spécialement adaptées à la gestion ou au fonctionnement d’équipement ou de dispositifs médicaux pour la gestion ou l’administration de ressources ou d’établissements de soins de santé, p.ex. pour la gestion du personnel hospitalier ou de salles d’opération

7.

DATA PREDICTION METHOD, APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020135601
Numéro de publication 2022/011946
Statut Délivré - en vigueur
Date de dépôt 2020-12-11
Date de publication 2022-01-20
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Yu, Woliang
  • Ma, Xiaozhen

Abrégé

A data prediction method, an apparatus, a computer device, and a storage medium, relating to the field of artificial intelligence, and the method comprising: on the basis of a data prediction request, determining model information and first user information, and acquiring a prediction data table; on the basis of the model data, acquiring a pre-generated data mining model from a model server, and on the basis of the first user information, allocating a corresponding prediction resource in the model server; and on the basis of the prediction resource and the data mining model, generating a prediction model file and sending same to a data storage server, in order to operate the data mining model on the data storage server, and on the basis of the prediction data table, acquiring corresponding data and inputting same into the data mining model to perform data prediction. In addition the method further relates to blockchain technology, and private information in the data acquired in the data mining model generation and data prediction processes may be stored in a blockchain. The present method is able to implement one-key generation of a data mining model and one-key data prediction deployment.

Classes IPC  ?

  • G06F 16/2458 - Types spéciaux de requêtes, p.ex. requêtes statistiques, requêtes floues ou requêtes distribuées

8.

METHOD AND DEVICE FOR TEXT-BASED IMAGE GENERATION

      
Numéro d'application CN2021103681
Numéro de publication 2022/007685
Statut Délivré - en vigueur
Date de dépôt 2021-06-30
Date de publication 2022-01-13
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Gou, Yuchuan
  • Wu, Qiancheng
  • Li, Minghao
  • Gong, Bo
  • Han, Mei

Abrégé

A method and device for image generation are provided. The method includes: obtaining a text describing a content of an image to be generated; extracting, using a text encoder, a text feature vector from the text; determining a semantic mask as spatial constraints of the image to be generated; and automatically generating the image using a generative adversarial network (GAN) model according to the semantic mask and the text feature vector.

Classes IPC  ?

  • G06T 11/00 - Génération d'images bidimensionnelles [2D]

9.

EMOTIONAL VOICE DATA CONVERSION METHOD, APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2021083126
Numéro de publication 2022/007438
Statut Délivré - en vigueur
Date de dépôt 2021-03-26
Date de publication 2022-01-13
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liu, Zhengchen
  • Chen, Minchuan
  • Ma, Jun
  • Wang, Shaojun
  • Xiao, Jing

Abrégé

An emotional voice data conversion method and a related device, able to be used in the fields of smart governance or smart medicine, and comprising: acquiring source emotional voice data, and on the basis of a pre-set vocoder, extracting a first fundamental frequency, non-periodic data, and a first Mel-frequency cepstrum coefficient from the source emotional voice data (S201); by means of a linear transform, converting the first fundamental frequency into a second fundamental frequency, and on the basis of a pre-set conversion model, converting the first Mel-frequency cepstrum coefficient into a second Mel-frequency cepstrum coefficient, the pre-set conversion model being of a variational autoencoder generative adversarial network structure; and on the basis of the vocoder, performing a Fourier transform and adaptive interpolation processing on the second fundamental frequency, the second Mel-frequency cepstrum coefficient, and the non-periodic data, to obtain target emotional data (S203).

Classes IPC  ?

  • G10L 19/18 - Vocodeurs utilisant des modes multiples
  • G10L 21/013 - Adaptation à la hauteur tonale ciblée
  • G10L 25/24 - Techniques d'analyses de la parole ou de la voix qui ne se limitent pas à un seul des groupes caractérisées par le type de paramètres extraits les paramètres extraits étant le cepstre

10.

PRE-LABELING METHOD BASED ON HIERARCHICAL TRANSFER LEARNING AND RELATED DEVICE

      
Numéro d'application CN2021082942
Numéro de publication 2022/001233
Statut Délivré - en vigueur
Date de dépôt 2021-03-25
Date de publication 2022-01-06
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhang, Nan
  • Wang, Jianzong
  • Qu, Xiaoyang

Abrégé

The embodiments of the present application relates to the field of artificial intelligence, and to a pre-labeling method based on hierarchical transfer learning and a related device, comprising: on the basis of a pre-set clustering algorithm, performing clustering on a plurality of different scenarios received in advance to obtain a clustering result; on the basis of the clustering result, determining a first classification of scenario and a second classification of scenario, the first classification of scenario comprising a first scenario, and the amount of labeled data in the first scenario being greater than the amount of labeled data in any scenario in the second classification of scenario; on the basis of a pre-set recognition model, performing transfer learning on the first classification of scenario to obtain pre-labeled data of each scenario in the first classification of scenario and a transfer model; and on the basis of the transfer model, performing transfer learning on the second classification of scenario to obtain pre-labeled data of each scenario in the second classification of scenario. The pre-labeled data in each scenario can be stored in a blockchain. The present application makes possible rapid acquisition of good pre-labeled data in different scenarios.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06N 20/00 - Apprentissage automatique

11.

METHOD AND APPARATUS FOR DETECTING PLURALITY OF TYPES OF SOUND EVENTS

      
Numéro d'application CN2021083752
Numéro de publication 2022/001245
Statut Délivré - en vigueur
Date de dépôt 2021-03-30
Date de publication 2022-01-06
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liu, Boqing
  • Wang, Jianzong
  • Zhang, Zhiyong
  • Cheng, Ning

Abrégé

A method of detecting a plurality of types of sound events, comprising: extracting a sound source matrix from sound source data (S100); inputting the sound source matrix into a trained feature extraction network to extract a feature matrix of a sound event (S200); inputting the feature matrix into a trained weight gating loop layer, and on the basis of weights of preceding vectors in a feature matrix and a weight matrix of the weight gating loop layer, weighting corresponding subsequent vectors in the feature matrix to obtain a weighted feature matrix (S300); inputting the weighted feature matrix into a fully connected layer, and by means of the full connection, acquiring a probability matrix, the number of dimensions of the probability matrix corresponding to the number of classifications of sound events (S400); on the basis of the probability matrix, and determining a target sound event that has occurred (S600). The sound source matrix can be stored in a blockchain.

Classes IPC  ?

  • G10L 17/02 - Opérations de prétraitement, p.ex. sélection de segment; Représentation ou modélisation de motifs, p.ex. fondée sur l’analyse linéaire discriminante [LDA] ou les composantes principales; Sélection ou extraction des caractéristiques
  • G10L 17/04 - Entraînement, enrôlement ou construction de modèle
  • G10L 17/18 - Réseaux neuronaux artificiels; Approches connexionnistes
  • G10L 17/22 - Procédures interactives; Interfaces homme-machine

12.

BLOCK CHAIN-BASED TRADE DATA PROCESSING METHOD AND RELATED DEVICE THEREOF

      
Numéro d'application CN2021096463
Numéro de publication 2022/001526
Statut Délivré - en vigueur
Date de dépôt 2021-05-27
Date de publication 2022-01-06
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Wang, Cong

Abrégé

Embodiments of the present application belong to the technical field of block chains, and relate to a block chain-based trade data processing method and a related device thereof. The method comprises: adding participants into a trade financing network by means of corresponding nodes; storing trade data uploaded by the participants into blocks, the participants comprising a financing party node and a fund party node; notifying the financing party node to specify an association relationship of the trade data when detecting that each participant completes uploading the trade data; when it is detected that the financing party node selects the trade data and initiates a financing application request on the basis of the selected trade data, determining the trade data having the association relationship with the selected trade data as data to be verified, and performing cross verification and evaluation calculation on the trade data selected by the financing party node and the data to be verified to obtain a verification result and an evaluation result, respectively; and sending the financing application request, the verification result and the evaluation result to the fund party node. According to the present application, the authenticity of data and the accuracy of a financing evaluation result are improved.

Classes IPC  ?

  • G06Q 20/06 - Circuits privés de paiement, p.ex. impliquant de la monnaie électronique utilisée uniquement entre les participants à un programme commun de paiement
  • G06Q 20/22 - Schémas ou modèles de paiement

13.

METHOD AND APPARATUS FOR QUESTION-AND-ANSWER DATA ENHANCEMENT, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2021082936
Numéro de publication 2022/001232
Statut Délivré - en vigueur
Date de dépôt 2021-03-25
Date de publication 2022-01-06
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Qiao, Yixuan
  • Chen, Hao
  • Gao, Peng

Abrégé

A method and an apparatus for question-and-answer data enhancement, a computer device, and a storage medium, relating to artificial intelligence technology, and specifically for use in deep learning. The method comprises acquiring a question and answer dataset, the question and answer dataset comprising a plurality of data points and real tags corresponding thereto (S1); on the basis of a pre-trained prediction model and the real tags, performing first soft tag prediction on each data point to obtain first soft tags corresponding to each data point (S2); constructing a soft tag dataset from each data point and the first soft tags corresponding thereto, and use knowledge distillation to generate a labeling model from the soft tag dataset and the prediction model (S3) acquiring a dataset to be tagged, inputting the dataset to be tagged into the labeling model for pre-labeling, and on the basis of labeling results, screening the dataset to be tagged to obtain a labeled sample set (S4). The described method further relates to blockchain technology, and the data in the labeled sample set and the dataset to be tagged are stored in a blockchain. The described method is able to improve efficiency and quality in labeling and tagging.

Classes IPC  ?

14.

BLOCKCHAIN LEDGER MANAGEMENT METHOD AND APPARATUS, AND COMPUTER DEVICE

      
Numéro d'application CN2021097547
Numéro de publication 2022/001549
Statut Délivré - en vigueur
Date de dépôt 2021-05-31
Date de publication 2022-01-06
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhang, Bao
  • Xie, Danli
  • Wang, Menghan
  • Liu, Enke
  • Cha, Luyan

Abrégé

The present application relates to the technical field of distributed ledgers of blockchains. Disclosed are a blockchain ledger management method and apparatus, and a computer device. The method comprises: receiving first transaction data written by a first user in an alliance chain network, wherein the first transaction data carries a synchronous ciphertext encrypted by the first user to the first transaction data; storing the first transaction data and the synchronous ciphertext of the first transaction data in a distributed ledger of the alliance chain network; associating, according to the first transaction data, a second user having a transaction relationship with the first user, wherein the second user and the first user are associated as a designated transaction pair; determining whether the first transaction and the second transaction in the designated transaction pair have been correctly executed; if yes, determining whether a confirmation result of a zero-knowledge proof of the transaction relationship of a third user to the first transaction pair in the alliance chain network is received; if yes, changing the transaction states of the first transaction and the second transaction in the distributed ledger to a confirmed state, and releasing same in the alliance chain network. The privacy and the security of the transactions are improved.

Classes IPC  ?

  • G06F 21/64 - Protection de l’intégrité des données, p.ex. par sommes de contrôle, certificats ou signatures

15.

METHOD FOR TASK ASSIGNMENT IN DISTRIBUTED COMPUTING, AND RELATED DEVICE

      
Numéro d'application CN2021096643
Numéro de publication 2021/258988
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Liu, Hongya

Abrégé

The present application relates to artificial intelligence and provides a method for task assignment in distributed computing and a related device. The method comprises: determining a weight of each compute node, and calculating a weight percentage of each compute node; constructing a neural network model comprising input layers, hidden layers and output layers, each input layer node and each hidden layer node corresponding to a computing task and each output layer node corresponding to a compute node; computing the sum of first weight-value parameters between the input layer nodes and the hidden layer nodes, and determining, according to the sum of the first weight-value parameters, second weight-value parameters between the hidden layer nodes and the output layer nodes; calculating a task assignment error of each compute node according to the second weight-value parameters; judging whether the task assignment error satisfies a convergence condition; adjusting the second weight-value parameters if the task assignment error does not satisfy the convergence condition; and modifying the first weight-value parameters. The present application can assign compute tasks evenly to each compute node.

Classes IPC  ?

  • G06F 9/48 - Lancement de programmes; Commutation de programmes, p.ex. par interruption

16.

OBJECT CONTOUR DELINEATION METHOD AND APPARATUS, AND COMPUTER SYSTEM AND READABLE STORAGE MEDIUM

      
Numéro d'application CN2021096666
Numéro de publication 2021/258991
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Li, Chengling
  • Xia, Minghao

Abrégé

An object contour delineation method and apparatus, and a computer system and a readable storage medium. The method comprises: receiving an image to be recognized, and calling a conversion tool of a front-end device to convert said image into a base 64 code to serve as an input parameter (S101); calling an image recognition system from a background interface, calculating the input parameter to obtain a recognition result, and sending the recognition result to the front-end device (S102); calling a contour conversion algorithm from the front-end device, and calculating contour coordinates to obtain contour positioning information (S103); and calling a cascading style sheet from the front-end device, and drawing a recognition box on said image according to the contour positioning information (S105). By means of the method, the currently existing problem of a system developer being unable to pinpoint the parts in an image which have been recognized by means of an image recognition system is solved, and a working process of the image recognition system is provided for the developer, such that the developer can optimize and adjust the image recognition system according to the working process.

Classes IPC  ?

  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales
  • G06K 9/32 - Alignement ou centrage du capteur d'image ou de la zone image

17.

INFORMATION PROPAGATION PATH ANALYSIS METHOD AND APPARATUS, AND COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2021096857
Numéro de publication 2021/258998
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Cao, Hexin
  • Cai, Jian

Abrégé

An information propagation path analysis method, comprising: acquiring propagation record information of a network (201); dividing the network into at least one community network according to the propagation record information, and determining candidate sampling nodes and an initial propagation node in each community network (202); calculating the node composition ratio of connection nodes, key nodes and ordinary nodes in the candidate sampling nodes by means of node type identifiers of the candidate sampling nodes (203); determining a sampling policy on the basis of the calculated node composition ratio and a preset node association relationship, wherein the node association relationship comprises the connection nodes being associated with the key nodes and the key nodes being associated with the ordinary nodes (204); sampling the initial propagation nodes and the candidate sampling nodes according to the sampling policy (205); and performing visual presentation on collected initial propagation nodes and candidate sampling nodes to generate an information propagation path map (206). In addition, the propagation record information can be stored in a blockchain. By means of the method, the accuracy of information propagation path analysis is improved.

Classes IPC  ?

  • G06F 16/9536 - Personnalisation de la recherche basée sur le filtrage social ou collaboratif

18.

DECISION TREE-BASED METHOD AND APPARATUS FOR OUTPUTTING ABNORMAL DATA SOURCES, AND COMPUTER DEVICE

      
Numéro d'application CN2021096979
Numéro de publication 2021/259002
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY(SHENZHEN)CO.,LTD. (Chine)
Inventeur(s)
  • Liu, Li
  • Liu, Zhongyuan
  • Zhao, Shiquan

Abrégé

The present application relates to the technical field of data mining of artificial intelligence. Disclosed are a decision tree-based method and apparatus for outputting abnormal data sources, a computer device, and a storage medium. The method comprises: obtaining an inputted data set; retrieving a preset knowledge graph; generating a knowledge node salvage tool; using the knowledge node salvage tool to correspondingly obtain a first node set and a second node set; obtaining a data stream set; obtaining a first data category sequence and a second data category sequence; counting the number of members of a specified data type; if the number of members is greater than a preset number threshold, obtaining abnormal members, and obtaining abnormal data streams; and recording sources at both ends of the abnormal data streams as the abnormal data sources, and outputting the abnormal data sources. Thus, the comprehensive and accurate discovery and output of the abnormal data sources are achieved. In addition, the present application further relates to a blockchain technology; and a decision tree can be stored in a blockchain.

Classes IPC  ?

  • G06F 16/36 - Création d’outils sémantiques, p.ex. ontologie ou thésaurus

19.

FEATURE RECOGNITION METHOD AND APPARATUS, AND COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2021096980
Numéro de publication 2021/259003
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Kong, Qingyang

Abrégé

The present application relates to a feature recognition method and apparatus, and a computer device and a storage medium. The method comprises: generating a plurality of candidate features from a plurality of original features according to a preset candidate feature generation method; generating a corresponding first meta-feature from each candidate feature according to a preset first meta-feature generation method; inputting the first meta-feature into a probability model, so as to obtain a target probability that each candidate feature is a preset label; comparing the target probability with a first preset threshold value, and all candidate features, the target probabilities of which are greater than or equal to the first preset threshold value, constituting a candidate feature set; combining the candidate features in the candidate feature set with the plurality of original features so as to calculate evaluation values of the candidate features; and comparing the evaluation values with a second preset threshold value, wherein candidate features, evaluation values corresponding to which are greater than the second preset threshold value, are effective features. By means of the feature recognition method and apparatus, and the computer device and the storage medium provided in the present application, features effective for machine learning are selected from constructed features.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques

20.

DATA DICTIONARY GENERATION METHOD AND APPARATUS, DATA QUERY METHOD AND APPARATUS, AND DEVICE AND MEDIUM

      
Numéro d'application CN2021090528
Numéro de publication 2021/258848
Statut Délivré - en vigueur
Date de dépôt 2021-04-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liu, Dongyv
  • Chen, Leqing
  • Zeng, Zengfeng
  • Li, Xuan

Abrégé

Disclosed are a data dictionary generation method and apparatus, and a computer device and a storage medium. The method comprises: acquiring first data to be stored, said data comprising a first pinyin node and a second pinyin node; performing a query in a preset first data dictionary on the basis of the first pinyin node and the second pinyin node, so as to determine a first index sequence and a second index sequence; processing the first index sequence and the second index sequence by using a CSR method, so as to obtain candidate index groups; querying a candidate frequency value of each candidate index group in a preset second data dictionary, and screening out, from the candidate index groups, a target index group, the candidate frequency value of which meets a preset requirement; and mapping and storing said data and the target index group to generate a third data dictionary. Recovery is performed by means of combining a first data dictionary and a second data dictionary to obtain a third data dictionary, thereby solving the problem of information redundancy during data storage.

Classes IPC  ?

21.

BUSINESS INDICATOR EVALUATION METHOD AND APPARATUS BASED ON ARTIFICIAL INTELLIGENCE, AND DEVICE AND MEDIUM

      
Numéro d'application CN2021090624
Numéro de publication 2021/258850
Statut Délivré - en vigueur
Date de dépôt 2021-04-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Cai, Jincheng

Abrégé

A business indicator evaluation method and apparatus based on artificial intelligence, and a device and a medium, which relate to the technical field of artificial intelligence and data processing. The method comprises: acquiring and determining a forward indicator and a reverse indicator in a system to be evaluated; calculating weights of the forward indicator and the reverse indicator by using an entropy method; calculating scores of the forward indicator and the reverse indicator on the basis of a P-IBI algorithm; calculating the weighted sum of the scores of the forward indicator and the reverse indicator; and performing standardization processing on the weighted sum to obtain a score of the system to be evaluated. The score of the system to be evaluated may be stored in a blockchain. According to the method, overall evaluation is performed by means of combining a weight that is determined by means of the entropy method and a score that is determined by means of the P-IBI algorithm, and the defect of indicator weights not being taken into consideration in the P-IBI algorithm is eliminated, such that an evaluation result is more in line with objective facts; moreover, the defect of poor readability of a score obtained by means of the P-IBI algorithm is overcome, and the automatic evaluation of business indicators is realized.

Classes IPC  ?

  • G06Q 10/06 - Ressources, gestion de tâches, gestion d'hommes ou de projets, p.ex. organisation, planification, ordonnancement ou affectation de ressources en temps, hommes ou machines; Planification dans l'entreprise; Modèles organisationnels

22.

VOCABULARY ERROR CORRECTION METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2021091066
Numéro de publication 2021/258853
Statut Délivré - en vigueur
Date de dépôt 2021-04-29
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Chen, Leqing
  • Liu, Dongyv
  • Zeng, Zengfeng
  • Li, Xuan

Abrégé

A vocabulary error correction method and apparatus, a computer device, and a storage medium. The method comprises: obtaining a vocabulary to be processed, said vocabulary comprising N candidate characters (S10); using an inverted index method to obtain, from a preset layered inverted index dictionary, candidate word sets corresponding to each candidate character to form a candidate word set collection, wherein the candidate word sets corresponding to each character in the layered inverted index dictionary are stored in a manner of classifying and layering according to the number of the characters (S20); using a Levenshtein distance algorithm to determine, from the vocabulary to be processed, a character to be replaced, and on the basis of the character to be replaced, determining, from the candidate word set collection, a word set to be processed, said word set comprising target characters and the candidate word sets corresponding to each target character (S30); and performing intersection processing on the candidate word sets corresponding to each target character to obtain a target vocabulary, wherein the target vocabulary is a vocabulary obtained after the vocabulary to be processed is subjected to the error correction (S40). The method solves the problem that the efficiency of vocabulary error correction is not high.

Classes IPC  ?

  • G06F 40/232 - Correction orthographique, p.ex. vérificateurs d’orthographe ou insertion des voyelles
  • G06F 40/284 - Analyse lexicale, p.ex. segmentation en unités ou cooccurrence

23.

DATA ENCRYPTION PROCESSING METHOD AND APPARATUS, AND COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2021091696
Numéro de publication 2021/258860
Statut Délivré - en vigueur
Date de dépôt 2021-04-30
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Yu, Yang
  • Ma, Ning

Abrégé

The present application relates to the technical field of data security in big data. Disclosed are a data encryption processing method and apparatus, and a computer device and a storage medium. The method comprises: screening, from source data, offset fields to be processed; acquiring user information corresponding to a specified offset field; acquiring a first data type of the specified offset field; according to the first data type, selecting a data transformation mode, which corresponds to the first data type, for the specified offset field; according to the data transformation mode, acquiring a plurality of pieces of encryption information, the number of which is the same as the number of all the pieces of user information; respectively performing corresponding data encryption transformation processing on the specified offset field by using each piece of encryption information, so as to obtain a plurality of processed specified offset fields, the number of which is the same as the number of pieces of encryption information; and sending, on one-to-one correspondence basis, the processed specified offset fields to user terminals respectively corresponding to the user information. By means of the present application, the data integrity and accuracy of an offset field during a data transmission process can be effectively guaranteed.

Classes IPC  ?

24.

USER BEHAVIOR MONITORING METHOD AND APPARATUS BASED ON BIG DATA, DEVICE, AND MEDIUM

      
Numéro d'application CN2021096700
Numéro de publication 2021/258992
Statut Délivré - en vigueur
Date de dépôt 2021-05-28
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Xu, Chaojun

Abrégé

A user behavior monitoring method based on big data. The method comprises: obtaining a plurality of users in a first preset time period and historical behavior records and identification information of the users (S101); taking the identification information as a target variable, taking the historical behavior records of the plurality of users as input parameters, and performing training by using a naive Bayesian algorithm to obtain an execution probability of each piece of identification information (S102); for a user to be monitored, taking identification information as a target variable, taking a historical behavior record of the user to be monitored as an input parameter, and using the naive Bayesian algorithm for prediction to obtain a user probability of the user to be monitored (S103); obtaining the identification information of the user to be monitored and an execution probability corresponding to the identification information (S104); and according to the user probability of the user to be monitored and the execution probability of the identification information, determining whether the user to be monitored has a risk (S105). The method solves the problem of user behavior monitoring caused by continuously increased data tables and cross access requirements of a background database. The method further relates to a blockchain and an artificial intelligence technology.

Classes IPC  ?

  • G06F 11/34 - Enregistrement ou évaluation statistique de l'activité du calculateur, p.ex. des interruptions ou des opérations d'entrée–sortie

25.

VIDEO-BASED MICRO-EXPRESSION RECOGNITION METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2021097208
Numéro de publication 2021/259005
Statut Délivré - en vigueur
Date de dépôt 2021-05-31
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Xiong, Wei

Abrégé

A video-based micro-expression recognition method and apparatus, a computer device, and a storage medium, relating to biological recognition of artificial intelligence. The method comprises: obtaining, in user video data, image frames containing micro-expressions; obtaining, from the image frames containing the micro-expressions, continuous multiple frame images having the number equal to the experience frame numerical value so as to form a micro-expression sequence; calling a weight calculation layer to calculate image feature vectors of combined weight values of frame images in the micro-expression sequence; summing the image feature vectors of combined weight values of the frame images to obtain a corresponding comprehensive image feature vector; inputting the comprehensive image feature vector into a convolutional neural network to obtain a micro-expression recognition result; and calling an item processing micro-expression strategy to obtain corresponding item processing flow information. According to the method, a time sequence relationship of the micro-expressions among a plurality of continuous image frames is fully considered when a neural network is used for classifying the micro-expressions, and time domain information of the micro-expressions in an image video sequence is learned, so that the micro-expression recognition result is obtained more accurately.

Classes IPC  ?

  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

26.

DATA PUSHING METHOD AND SYSTEM BASED ON BEHAVIOR TRAJECTORY, AND COMPUTER DEVICE

      
Numéro d'application CN2021097278
Numéro de publication 2021/259007
Statut Délivré - en vigueur
Date de dépôt 2021-05-31
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Zhang, Jie

Abrégé

A data pushing method based on a behavior trajectory. The method comprises: receiving a data pushing instruction triggered by a target user by means of a target user terminal (S100); according to the data pushing instruction, acquiring target historical behavior data of the target user from a behavior log system (S102); constructing a target trajectory graph of the target user according to the target historical behavior data (S104); extracting a target trajectory feature vector of the target user according to the target trajectory graph (S106); inputting the target trajectory feature vector into a recall model so as to obtain a plurality of pieces of initial pushing data (S108); inputting the plurality of pieces of initial pushing data into a classification model so as to obtain a plurality of pieces of target pushing data (S110, S112); and pushing the plurality of pieces of target pushing data to a front end for presentation (S114). In the method, a trajectory graph is constructed for a user according to behavior trajectories of the user, and similarity matching is performed on the trajectories of the user by means of a classification model and a regression model, thereby improving the accuracy and efficiency of data recommendation.

Classes IPC  ?

  • G06F 16/9536 - Personnalisation de la recherche basée sur le filtrage social ou collaboratif

27.

METHOD AND APPARATUS FOR CONSTRUCTING FACIAL ACTION UNIT DATA SET, AND COMPUTER DEVICE

      
Numéro d'application CN2021097549
Numéro de publication 2021/259012
Statut Délivré - en vigueur
Date de dépôt 2021-05-31
Date de publication 2021-12-30
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Luo, Linyao
  • Xu, Guoqiang

Abrégé

The present invention relates to data processing. Provided are a method and apparatus for constructing a facial action unit data set, and a computer device. The method comprises: acquiring facial pictures, and generating a plurality of triples on the basis of the facial pictures; pre-processing the facial pictures in the triples, and then inputting same into a network model, so as to obtain a feature vector corresponding to each triple; calculating the feature vector to obtain a loss value, calculating a parameter gradient by means of network back propagation, and updating a model parameter until the model converges, so as to obtain a search model; acquiring category pictures of various different facial action categories, and inputting each category picture into the search model; performing calculation by means of the search model, so as to obtain a facial feature of each category picture, and a Euclidean distance between the facial feature of each category picture and a facial feature in a specified database, and acquiring a pre-set number of facial pictures according to each Euclidean distance, so as to form a picture set; and labeling each picture set and generating a facial action unit data set. By means of a retrieval model, a great number of facial pictures of different categories can be quickly acquired, thereby enriching the types and the quantity of samples of a data set.

Classes IPC  ?

  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales

28.

INTELLIGENT EXERCISE RECOMMENDATION METHOD AND APPARATUS, COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020098828
Numéro de publication 2021/253480
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-12-23
Propriétaire PING AN TECHNOLOGY(SHENZHEN)CO., LTD. (Chine)
Inventeur(s)
  • Liang, Jin
  • Feng, Xin
  • Sheng, Liang
  • Xie, Baolin

Abrégé

An intelligent exercise recommendation method and apparatus, a computer device and a storage medium, which relate to the field of artificial intelligence. The method comprises: by means of a first preset formula, according to an exercise completion situation of each student, first calculating knowledge points which are fully mastered by each student and knowledge points which are possibly not fully mastered; and then by means of a second preset formula, further calculating exercises which are appropriate to be recommended to the students. The problems of wrong exercises being relatively singular in existing wrong exercise recommendation and the quality pertinence of wrong exercise recommendation not being high are solved. Personalized recommended practice exercises can be automatically generated with regard to the work situations and knowledge point mastering situations of students, so that the students spend time on exercises which need to be practiced the most, thereby helping the students to learn efficiently, and the exercise completion time can thus be reduced, thereby achieving the aims of performing personalized learning and improving the learning efficiency.

Classes IPC  ?

  • G06F 16/335 - Filtrage basé sur des données supplémentaires, p.ex. sur des profils d’utilisateurs ou de groupes

29.

DEVICE AND METHOD FOR ALIGNMENT OF MULTI-MODAL CLINICAL IMAGES USING JOINT SYNTHESIS, SEGMENTATION, AND REGISTRATION

      
Numéro d'application CN2021094548
Numéro de publication 2021/238732
Statut Délivré - en vigueur
Date de dépôt 2021-05-19
Date de publication 2021-12-02
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liu, Fengze
  • Cai, Jinzheng
  • Huo, Yuankai
  • Lu, Le
  • Harrison, Adam, P.

Abrégé

An image processing method for performing image alignment includes: acquiring a moving image generated by a first imaging modality; acquiring a fixed image generated by a second imaging modality; jointly optimizing a generator model, a register model, and a segmentor model applied to the moving image and the fixed image according to a plurality of cost functions; and applying a spatial transformation corresponding to the optimized register model to the moving image to align the moving image to the fixed image; wherein: the generator model generates a synthesized image from the moving image conditioned on the fixed image; the register model estimates the spatial transformation to align the synthesized image to the fixed image; and the segmentor model estimates segmentation maps of the moving image, the fixed image, and the synthesized image.

Classes IPC  ?

  • G06T 7/33 - Détermination des paramètres de transformation pour l'alignement des images, c. à d. recalage des images utilisant des procédés basés sur les caractéristiques
  • G06T 7/00 - Analyse d'image

30.

IDENTITY ENCRYPTION-BASED DATA TRANSMISSION METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2021096243
Numéro de publication 2021/239034
Statut Délivré - en vigueur
Date de dépôt 2021-05-27
Date de publication 2021-12-02
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Xie, Danli
  • Jia, Mu
  • Zhang, Pengcheng
  • Lu, Frank Yifan Chen

Abrégé

The present application relates to the blockchain technology, and provides an identity encryption-based data transmission method, comprising: allocating a first public/private key pair and a second public/private key pair to a user B (S1); randomly allocating a third private key pair to a user A (S2); calculating a public key address of the second public/private key pair according to the first public/private key pair, the second public/private key pair and a third public/private key (S3); encrypting a dataset according to the public key address to obtain an encrypted dataset, and using the encrypted dataset to calculate a transmission public key address transmitted to the user B (S4); verifying whether the public key address is the same as the transmission public key address or not (S5); and if the public key address is the same as the transmission public key address, allowing the dataset to be transmitted to the user B (S7). The present application also provides an identity encryption-based data transmission apparatus, an electronic device and a computer readable storage medium. The present application carries out relationship calculation according to the public/private key pairs of the user, so that the data transmission security is improved.

Classes IPC  ?

  • H04L 29/06 - Commande de la communication; Traitement de la communication caractérisés par un protocole

31.

MULTI-ROUND DIALOGUE MANAGEMENT METHOD AND APPARATUS, AND COMPUTER DEVICE

      
Numéro d'application CN2021096514
Numéro de publication 2021/239069
Statut Délivré - en vigueur
Date de dépôt 2021-05-27
Date de publication 2021-12-02
Propriétaire PING AN TECHNOLOGY(SHENZHEN)CO., LTD. (Chine)
Inventeur(s)
  • Liu, Minghui
  • Xu, Guoqiang

Abrégé

The present application discloses a multi-round dialogue management method, comprising: receiving an inputted first dialogue, and extracting structural information of the first dialogue; according to the structural information of the first dialogue and an optimization argument pre-learned by a neural network model in a dialogue management module, determining inference parameter information Q(S, A; θ) matching the first dialogue, wherein θ is the optimization argument obtained by training for optimization, the optimization argument is obtained by training according to preset expert knowledge, and the expert knowledge comprises an optimal dialogue decision action item in a preset dialogue state, S represents a dialogue state, and A represents a reward distribution of the dialogue decision action item corresponding to the dialogue state; according to the inference parameter information, selecting a first dialogue decision action item matching the structural information of the first dialogue; and controlling the first dialogue decision action item to decide a second dialogue matching the structural information of the first dialogue, and outputting the second dialogue. A convergence speed and an accuracy rate of a training process are increased.

Classes IPC  ?

32.

ENTERPRISE OPERATION DATA ANALYSIS METHOD AND APPARATUS BASED ON CONFIGURATION ALGORITHM, AND ELECTRONIC DEVICE AND MEDIUM

      
Numéro d'application CN2021090719
Numéro de publication 2021/238563
Statut Délivré - en vigueur
Date de dépôt 2021-04-28
Date de publication 2021-12-02
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Chen, Xitao

Abrégé

Provided are an enterprise operation data analysis method based on a configuration algorithm. The method comprises: acquiring an algorithm configuration main table, an algorithm configuration parameter table and an algorithm parameter value table; selecting an algorithm from the algorithm configuration main table, and calculating the selected algorithm, so as to obtain an output variable set; performing estimation calculation on the output variable set, so as to obtain a core correlation value; calculating an error value for the core correlation value, and when the error value is less than or equal to a preset error threshold value, using a parameter and a parameter value to configure the selected algorithm; and acquiring enterprise data, and using the algorithm, the configuration of which is completed, to perform data analysis on the enterprise data, so as to obtain an operation condition of an enterprise. In addition, blockchain technology is further involved. An algorithm configuration main body, an algorithm configuration parameter table and an algorithm parameter value table are stored in a blockchain. The problems of the high cost and low efficiency of data analysis operated by an enterprise can thus be solved.

Classes IPC  ?

  • G06F 16/22 - Indexation; Structures de données à cet effet; Structures de stockage
  • G06F 16/25 - Systèmes d’intégration ou d’interfaçage impliquant les systèmes de gestion de bases de données

33.

ABNORMAL COMMUNITY DETECTION METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2021096155
Numéro de publication 2021/239004
Statut Délivré - en vigueur
Date de dépôt 2021-05-26
Date de publication 2021-12-02
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Cao, Hexin

Abrégé

An abnormal community detection method comprises: constructing a guarantee relationship network, and segmenting the guarantee relationship network so as to obtain communities having abnormal guarantee relationships (S200); determining feature information of each of the communities, wherein the feature information comprises at least one of the following: a node scale, an edge scale, a clustering coefficient, the number of connected triangles, and an average degree (S300); determining, according to the feature information, communities having similar features to be a relationship cluster (S400); calculating the respective Euclidean distances of relationship clusters (S500); classifying the relationship clusters according to the Euclidean distances, and determining whether or not a relationship cluster is an abnormal cluster on the basis of a classification result (S600); and if so, determining that communities in the abnormal cluster are abnormal communities, and extracting the abnormal communities (S700). The method achieves efficient extraction of abnormal communities. In addition, the method relates to blockchain technology as the feature information can be stored in a blockchain.

Classes IPC  ?

  • G06Q 50/00 - Systèmes ou procédés spécialement adaptés à un secteur particulier d’activité économique, p.ex. aux services d’utilité publique ou au tourisme
  • G06Q 40/08 - Assurance, p.ex. analyse des risques ou pensions

34.

KEY ROTATION METHOD, DEVICE, ELECTRONIC APPARATUS, AND MEDIUM

      
Numéro d'application CN2021096434
Numéro de publication 2021/239059
Statut Délivré - en vigueur
Date de dépôt 2021-05-27
Date de publication 2021-12-02
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Shen, Xiangwen

Abrégé

The present application pertains to information security, and provides a key rotation method. The method comprises: generating multiple key pairs, and establishing an index for the multiple key pairs; extracting a random key pair from the multiple key pairs, and acquiring all stock data from a configuration database; performing, by using a random public key of the random key pair, encryption computation on the entire stock data to obtain first ciphertext corresponding to the entire stock data; determining, upon receiving a rotation request, a private key corresponding to the public key according to the index; performing, by using the private key, decryption computation on the first ciphertext to obtain plaintext; extracting a target key pair from the multiple key pairs; and performing, by using a target public key in the target key pair, encryption computation on the plaintext, and generating second ciphertext to respond to the rotation request, thereby improving a security level of all of the stock data. The present application further relates to block chain technology, and the target key pair is stored in a block chain.

Classes IPC  ?

  • H04L 9/30 - Clé publique, c. à d. l'algorithme de chiffrement étant impossible à inverser par ordinateur et les clés de chiffrement des utilisateurs n'exigeant pas le secret
  • H04L 9/08 - Répartition de clés
  • H04L 29/06 - Commande de la communication; Traitement de la communication caractérisés par un protocole
  • G06Q 20/38 - Architectures, schémas ou protocoles de paiement - leurs détails
  • G06Q 40/04 - Opérations boursières, p.ex. actions, marchandises, produits dérivés ou change de devises
  • G06F 21/64 - Protection de l’intégrité des données, p.ex. par sommes de contrôle, certificats ou signatures
  • G06F 16/22 - Indexation; Structures de données à cet effet; Structures de stockage

35.

RESPONSE METHOD AND APPARATUS BASED ON EMOTION RECOGNITION, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020092977
Numéro de publication 2021/217769
Statut Délivré - en vigueur
Date de dépôt 2020-05-28
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Ye, Yizhou
  • Hu, Hongwei
  • Ma, Jun
  • Wang, Shaojun

Abrégé

A response method and apparatus based on emotion recognition, a computer device, and a storage medium. The method comprises: obtaining target text information corresponding to information to be recognized from a user terminal; obtaining, according to a text processing rule, a text feature vector corresponding to the target text information; obtaining, according to an emotion recognition model, an emotion level corresponding to the text feature vector; and obtaining corpus information, which matches the emotion level and the target text information, from a corpus information database as response corpus information, and feeding back same to the user terminal to complete response. On the basis of emotion recognition technology, information to be recognized from a user terminal can be received and an emotion level matching the information to be recognized can be obtained, and response corpus information can be obtained according to the emotion level and target text information corresponding to the information to be recognized; thus, response corpus information can be flexibly adjusted according to the emotional level in information to be recognized, thereby improving the flexibility of giving response to question information.

Classes IPC  ?

  • G10L 25/63 - Techniques d'analyses de la parole ou de la voix qui ne se limitent pas à un seul des groupes spécialement adaptées pour un usage particulier pour comparaison ou différentiation pour estimer un état émotionnel

36.

ENTERPRISE PUBLIC OPINION ANALYSIS METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND MEDIUM

      
Numéro d'application CN2020098236
Numéro de publication 2021/217843
Statut Délivré - en vigueur
Date de dépôt 2020-06-24
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhou, Xinxin
  • Li, Wen
  • Wu, Xiaobei
  • Yu, Xiuming
  • Wang, Wei

Abrégé

An enterprise public opinion analysis method and apparatus, an electronic device, and a storage medium, which are applied to the field of financial technology. The method comprises: acquiring public opinion texts; pre-processing the public opinion texts to obtain a plurality of word groups; determining, to be target texts, public opinion texts in which an enterprise entity is identified; carrying out subject identification on the target texts to obtain an identification result, and classifying the target texts to obtain a classification result; inputting the target texts, the identification result and the classification result into a pre-trained emotion model, so as to obtain an emotion result; selecting a target text where the emotion result of which is a negative sentiment, and taking same as an early-warning text; matching the early-warning text with early-warning word groups in a configuration list, and determining a matched early-warning word group to be the early-warning category of the early-warning text; and acquiring a target enterprise of the early-warning text, and generating early-warning information according to the early-warning category and the early-warning text. By means of the method, an early warning can be sent out in a timely and accurate manner.

Classes IPC  ?

  • G06F 16/9535 - Adaptation de la recherche basée sur les profils des utilisateurs et la personnalisation

37.

HUMAN BODY FALLING DETECTION METHOD AND APPARATUS, AND COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020098817
Numéro de publication 2021/217848
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Wang, Lihua
  • Gong, Lianyin
  • Zhou, Bao
  • Chen, Yuanxu

Abrégé

A human body falling detection method and apparatus, a computer device, and a storage medium. The method comprises carrying out three instances of determination, wherein the first instance of determination involves determining whether the ratio of the length of a first minimum rectangular frame to the width thereof meets the condition of being less than or equal to 1; if so, carrying out the second instance of determination, which involves determining whether the central point coordinates (x1, y1) of the first minimum rectangular frame is higher than the central point coordinates (x3, y3) of a second minimum rectangular frame of a physical device; and if not, carrying out the third instance of determination, which involves determining whether the posture of a human body conforms to a preset falling posture, etc. According to the method, by means of the three instances of determination, the determination algorithm is simple, and the accuracy of determination results is relatively high.

Classes IPC  ?

  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales

38.

BLOCKCHAIN NODE SYNCHRONIZATION METHOD, APPARATUS AND DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020098929
Numéro de publication 2021/217849
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Lu, Frank Yifan Chen
  • Yao, Jie

Abrégé

The present application relates to blockchain technology, and provides a blockchain node synchronization method, apparatus and device, and a storage medium. The method comprises: selecting, on the basis of a preset election algorithm, a master node from nodes in the same organization; performing, by means of the master node, transaction processing on transactions to be verified in a transaction block, performing, on the basis of the transaction processing result corresponding to said transactions, validity tagging on said transactions, and generating transaction information corresponding to said transactions; and sending the transaction information after the processing and the transaction block to slave nodes, and performing, by the slave nodes, transaction information synchronization on the basis of the verification result of the master node.

Classes IPC  ?

  • H04L 29/08 - Procédure de commande de la transmission, p.ex. procédure de commande du niveau de la liaison
  • G06F 16/27 - Réplication, distribution ou synchronisation de données entre bases de données ou dans un système de bases de données distribuées; Architectures de systèmes de bases de données distribuées à cet effet

39.

ABNORMAL CELL AUTOMATIC LABELING METHOD AND APPARATUS, ELECTRONIC DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020098969
Numéro de publication 2021/217851
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO.,LTD. (Chine)
Inventeur(s)
  • Guo, Bingxue
  • Wang, Jiyong
  • Chu, Xiao
  • Wang, Jian
  • Ping, Bo
  • Yu, Lin

Abrégé

The present application relates to an image processing technique, being applied to the field of smart medical care. Disclosed is an abnormal cell automatic labeling method, comprising: performing Gaussian convolution smoothing processing on a cell pathology picture to obtain an image pyramid; performing detection and fitting on the image pyramid by means of a Hough transform circle detection method so as to obtain a low-power fitting region of interest; mapping the low-power fitting region of interest to an obtained cell pathology picture so as to generate a fitting high-power image; segmenting a region of interest in the fitting high-power image to generate a segmented high-power image; labeling and cutting the segmented high-power image by means of an adaptive threshold segmentation algorithm to obtain an abnormal cell labeling set; and mapping the abnormal cell labeling set to the cell pathology picture to obtain an abnormal cell labeling picture. The present application improves the accuracy of abnormal cell labeling, and reduces the calculation and storage pressure. In addition, also involved is a blockchain technique, and the abnormal cell labeling set may be stored in a blockchain.

Classes IPC  ?

40.

DAMAGE DETECTION METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND MEDIUM

      
Numéro d'application CN2020098971
Numéro de publication 2021/217852
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhao, Xiaohong
  • Liu, Lihong
  • Liu, Yuyu

Abrégé

Disclosed is a damage detection method, comprising: acquiring an original damaged image set and denoising the original damaged image set to obtain a standard image set (S1 to S2); performing feature processing on the standard image set to obtain a feature image set (S3); training a plurality of image generation models by using the feature image set, and inputting the original damaged image set into the image generation models to obtain a damaged image training set and a damage feature marking image training set (S4 to S6); training a target part damage detection model by using the damaged image training set and the damage feature marking image training set (S7); and detecting, by using the target part damage detection model, a damaged image to be subjected to detection, so as to obtain a damage type (S8). Data for model training and detection can be stored in a blockchain. Disclosed are a damage detection apparatus, an electronic device, and a computer-readable storage medium for solving the problem of inaccurate model training caused due to damage data of a target part being insufficient.

Classes IPC  ?

  • G06K 9/32 - Alignement ou centrage du capteur d'image ou de la zone image
  • G06K 9/46 - Extraction d'éléments ou de caractéristiques de l'image
  • G06T 5/00 - Amélioration ou restauration d'image
  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion
  • G06N 3/08 - Méthodes d'apprentissage
  • G06Q 40/08 - Assurance, p.ex. analyse des risques ou pensions

41.

INTELLIGENT LOSS ASSESSMENT METHOD AND APPARATUS FOR DAMAGE IMAGE, ELECTRONIC DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020098972
Numéro de publication 2021/217853
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhao, Xiaohong
  • Liu, Lihong
  • Liu, Yuyu

Abrégé

An intelligent loss assessment method for a damage image, comprising: obtaining an original damage image set and sampling same according to a historical damage image set to obtain a sampling image set (S1); obtaining all the feature layers of each sampling image in the sampling image set, and adjusting the size of each feature layer in the sampling image as the same size to obtain an initial damage image set (S2); calculating semantic information of each feature layer in the initial damage image set, and using a preset semantic feature reinforcement function to perform feature reinforcement on the semantic information to obtain a standard damage image set (S3); calculating a damage value of each feature layer in the standard damage image set, detecting the damage position of the standard damage image set according to the damage value, and cropping images in the standard damage image set corresponding to the damage position to obtain a target damage image set (S4); and performing damage degree detection on the target damage image set by means of an image damage degree detection model completed pre-training to obtain a damage degree of the target damage image set (S5). The method also relates to a blockchain technology. Private information of a user can be stored in a blockchain node. The method can improve the loss assessment efficiency and the detection efficiency of the damage image.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

42.

FALSE POSITIVE FILTERING METHOD, DEVICE, EQUIPMENT, AND STORAGE MEDIUM

      
Numéro d'application CN2020098974
Numéro de publication 2021/217854
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Chen, Kaixing
  • Zhou, Xin
  • Lv, Chuanfeng

Abrégé

Related to the technical field of artificial intelligence, disclosed is a false positive filtering method. The method comprises: processing an image to be tested by means of a deep neural network model to position a normal area and a suspected area (S400); employing a preset rule to determine an initial normal area in the normal area outputted by the model (S402); employing a similarity measurement algorithm to calculate respectively an intraclass distance of the initial normal area and an interclass distance between the suspected area and the initial normal area (S404); calculating, on the basis of the intraclass distance and the interclass distance, the probability of the suspected area being a normal area (S406); and filtering out a false positive area on the basis of the calculated probability and of a selected threshold (S408). Also provided are a device, equipment, and a storage medium. This allows intraclass and interclass differences of data itself to be utilized, by comparing the degree of similarity between a normal area and a suspected in a same image to be tested, to distinguish true and false positive areas, thus supplementing and optimizing a network learning scheme.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques

43.

METHOD AND APPARATUS FOR LOCATING ROOT CAUSE OF ALARM, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020099434
Numéro de publication 2021/217865
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Chen, Zhenbo
  • Jin, Ge
  • Xu, Liang

Abrégé

A method for locating the root cause of an alarm, and a related apparatus, which relate to artificial intelligence. The method comprises: acquiring a particular alarm object in an alarm slice and generating a corresponding alarm cluster; aggregating monitoring indexes in specified alarm objects in the alarm cluster and generating an entry index; acquiring first index time sequence data corresponding to the specified alarm objects and acquiring second index time sequence data corresponding to the specified alarm objects; adjusting a first time window of the first index time sequence data according to a second time window of the second index time sequence data, and calculating a specified Pearson similarity between a specified index and the entry index; according to the specified Pearson similarity corresponding to each specified alarm object and the specified time difference corresponding to each specified Pearson similarity, selecting at least one specified alarm object from among all of the specified alarm objects as a root cause object of the alarm cluster and outputting same. The described method can quickly generate a root cause object related to an alarm object.

Classes IPC  ?

  • H04L 12/24 - Dispositions pour la maintenance ou la gestion
  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques

44.

METHOD AND APPARATUS FOR AI INTERVIEW RECOGNITION, COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020099520
Numéro de publication 2021/217866
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Deng, Yue
  • Jin, Ge
  • Xu, Liang

Abrégé

A method and an apparatus for artificial intelligence (AI) interview recognition, a computer device and a storage medium, relating to the field of AI-based semantic recognition. Said method comprises: acquiring text data; performing, sentence by sentence, vector segmentation on the text data, and determining at least one group of first text vectors and second text vectors corresponding to the first text vectors; importing the at least one group of the first text vectors and the second text vectors into a preset characterization model for processing, so as to obtain influence probability data of the several first text vectors and the second text vectors (S300); and ranking the influence probability data in a descending order from highest to lowest, and returning candidates corresponding to a preset number of pieces of top ranked influence probability data as a recognition result of an intelligent interview (S400), and storing the result in a blockchain network. Said method is able to effectively complete the recognition of an interview result, with high recognition precision.

Classes IPC  ?

  • G06F 9/48 - Lancement de programmes; Commutation de programmes, p.ex. par interruption

45.

XGBOOST-BASED DATA CLASSIFICATION METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020099521
Numéro de publication 2021/217867
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Li, Heng
  • Yuan, Dongsheng
  • Chen, Youxin
  • Xu, Liang

Abrégé

An XGBoost-based data classification method and apparatus, a computer device, and a storage medium. The method comprises: extracting, from a survey database, survey data corresponding to a disease to be classified as initial data (S1); using an XGBoost model to extract features from the initial data to obtain target features of the survey data (S2); using the target features to train the XGBoost model, and using a Hyperopt database to adjust model parameters in the XGBoost model to obtain a trained classification model (S3); generating a classification screening questionnaire according to the target features (S4); and for any user, taking data obtained from user's answers to the classification screening questionnaire as target data of a target user, and inputting the target data into the trained classification model, and outputting a category corresponding to the target user (S5). The method also involves blockchain technology, and the target data is stored in a blockchain. The method can achieve data analysis of different physiological features of different diseases, thereby improving accuracy of disease data classification.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques

46.

CLUSTER INSTALLATION METHOD, DEVICE, AND APPARATUS USING EXCEL FILE, AND STORAGE MEDIUM

      
Numéro d'application CN2020099542
Numéro de publication 2021/217870
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Lin, Qichen
  • Zhang, Guancheng
  • Wan, Shuwu
  • Li, Jun

Abrégé

A cluster installation method, device, and apparatus using an Excel file, and a storage medium, relating to the technical field of cloud services, and capable of reducing memory consumption and CPU resource consumption and improving the efficiency of cluster deployment. The cluster installation method using an Excel file comprises: receiving an Excel spreadsheet file uploaded by a terminal, the Excel file comprising addresses of main machines to be installed and information of components to be installed (101); analyzing and verifying the Excel file to obtain data content of the Excel file (102); counting the number of the main machines according to the data content of the Excel file, and generating, according to the number of the main machines, an order to be paid (103); sending the order to the terminal to instruct a target user to pay the order (104); and upon detecting that the target user has successfully paid the order, performing cluster deployment by means of a preset cloud service and the data content of the Excel file (105).

Classes IPC  ?

47.

EMOTION INFORMATION RECOGNITION METHOD AND APPARATUS, AND STORAGE MEDIUM AND COMPUTER DEVICE

      
Numéro d'application CN2020111036
Numéro de publication 2021/217973
Statut Délivré - en vigueur
Date de dépôt 2020-08-25
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Yu, Lingwei
  • Zhou, Chen
  • Zhou, Bao
  • Chen, Yuanxu

Abrégé

Disclosed are an emotion information recognition method and apparatus, and a storage medium and a computer device, which relate to the technical field of artificial intelligence, and aim to respectively process a human body posture and facial expression, which have been obtained, into a posture matrix and emotion vector information, and to simultaneously process, by means of a pre-trained emotion intensity recognition model, the posture matrix and the vector information to obtain emotion intensity data and correspondingly recognize an emotion type, thereby improving the accuracy and efficiency of emotion information recognition. The method comprises: receiving an emotion information recognition request, wherein the emotion information recognition request carries human body posture information; transforming the human body posture information into a posture matrix including posture feature points by using a preset posture transformation algorithm; processing the posture matrix according to a preset emotion intensity algorithm, so as to obtain emotion intensity data; and retrieving and feeding back a corresponding emotion type according to the emotion intensity data. In addition, the present application further relates to blockchain technology, and the emotion intensity data may be stored in a blockchain.

Classes IPC  ?

  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales

48.

SPOKEN LANGUAGE SEMANTIC MATCHING METHOD AND DEVICE

      
Numéro d'application CN2020111070
Numéro de publication 2021/217974
Statut Délivré - en vigueur
Date de dépôt 2020-08-25
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Wu, Xinchao
  • Zhou, Chen
  • Luo, Jiawei
  • Chen, Yuanxu

Abrégé

A spoken language semantic matching method and device, relating to the technical field of data processing. The method comprises: obtaining matching texts in a matching database, and splicing a text to be matched and the matching texts to generate spliced texts (101); computing first model codes of the spliced texts according to a first preset model, and computing second model codes of the spliced texts according to a second preset model (102); respectively extracting, according to a mask weight, first similarity dimensions of the first model codes and second similarity dimensions of the second model codes (103); splicing the first similarity dimensions and the second similarity dimensions to obtain splicing results, and extracting a maximum dimension value from the splicing results (104); and determining the matching text corresponding to the maximum dimension value as a semantically similar text of the text to be matched (105). The method is mainly applied in a human-machine interaction process.

Classes IPC  ?

49.

COOPERATIVE CONTROL METHOD AND APPARATUS FOR MULTIPLE ROBOTS, AND COMPUTER DEVICE

      
Numéro d'application CN2020111369
Numéro de publication 2021/217977
Statut Délivré - en vigueur
Date de dépôt 2020-08-26
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Yu, Lingwei
  • Fu, Zuoyi
  • Zhou, Chen
  • Zhou, Bao

Abrégé

Disclosed are a cooperative control method and apparatus for multiple robots, and a computer device, which relate to the technical field of computers. The method comprises: first, acquiring object information of an object to be transported and robot information of at least one currently available robot; identifying, according to the object information, information of the position of a transport insertion hole of said object; according to the position information and the robot information, respectively sending a control instruction to the at least one robot, so as to control front connectors of the at least one robot to be inserted into respective transport insertion holes one by one; and if it is determined that the at least one robot has been inserted into the respective transport insertion hole, respectively sending, to the at least one robot, an instruction for indicating transporting. The present application can realize accurate cooperative control of multiple robots in place of a manual control method, and reduce control deviations.

Classes IPC  ?

50.

DISTRIBUTED MESSAGE TRANSMISSION METHOD AND APPARATUS, COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020112347
Numéro de publication 2021/217986
Statut Délivré - en vigueur
Date de dépôt 2020-08-31
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO.,LTD. (Chine)
Inventeur(s) Zhou, Jian

Abrégé

Disclosed in the present application are a distributed message transmission method and apparatus, a computer device and a storage medium. The method comprises: a client sends a service message to first middleware, stores the service message in a customer database and returns a storage result; if the storage succeeds, the client executes a corresponding target service and sends an execution result to the first middleware; if the execution succeeds, the first middleware updates the state of the service message to "confirmed" and sends the service message to a server by means of an MQ; after receiving the service message, the server returns a service confirmation message to the client by means of the MQ. According to the preset application, a step of processing the message by the first middleware is added, and by skillfully processing the execution, storage and maintenance of the service message in the same transaction, the possibility of failure to send the service message to the MQ is decreased, the problem of transaction interruption caused by a service message transmission fault is reduced, and no manual intervention is required. Therefore, the method is convenient and efficient, and improves the service message transmission efficiency and the message data consistency.

Classes IPC  ?

51.

METHOD AND DEVICE FOR GENERATING SIMILAR TEXT

      
Numéro d'application CN2020117946
Numéro de publication 2021/218015
Statut Délivré - en vigueur
Date de dépôt 2020-09-25
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Luo, Jiawei
  • Wu, Xinchao
  • Gong, Lianyin
  • Zhou, Bao
  • Chen, Yuanxu

Abrégé

A method and a device for generating a similar text, relating to the field of semantic parsing technology. Said method mainly comprises: acquiring text-segmented words of an initial text (101); searching for text word vectors of the text-segmented words according to a preset word vector algorithm (102); splicing the text word vectors and relative position vectors of the text word vectors to generate spliced vectors (103); inputting the spliced vectors into a preset encoder to generate a set of characteristic word vectors of the initial text (104); and inputting the set of characteristic word vectors into a preset decoder to resolve a similar text of the initial text (105). Said method is mainly applied to a natural language processing process, and also relates to a blockchain technique. Spliced vectors can be stored in a blockchain node.

Classes IPC  ?

52.

METHOD AND APPARATUS FOR IMPLEMENTING MOUSE CURSOR TRAILING SPECIAL EFFECT, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020118002
Numéro de publication 2021/218017
Statut Délivré - en vigueur
Date de dépôt 2020-09-27
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Li, Qingpeng

Abrégé

Disclosed are a method and apparatus for implementing a mouse cursor trailing special effect, a computer device, and a storage medium, applied to the technical field of computers and used for achieving a linear mouse cursor special effect at a web page end. The method provided in the present application comprises: when a mouse cursor operation event is detected, obtaining coordinate points operated by a mouse cursor in a browser page; sequentially storing the obtained coordinate points in an array according to the chronological order of the event; obtaining, in real time, coordinate points through which the mouse cursor passes in the browser page, and updating and storing the coordinate points in the array; sequentially connecting every two adjacent coordinate points stored in the array into line segments to form a motion trail of the mouse cursor, and displaying the motion trail in a pre-created canvas; and every time a browser renders a page frame, erasing the motion trail displayed in the canvas, obtaining most recently updated coordinate points of the mouse cursor in the array, and recurrently implementing the steps for forming the motion trail of the mouse cursor and this step until the mouse cursor operation event is stopped.

Classes IPC  ?

  • G06F 3/0354 - Dispositifs de pointage déplacés ou positionnés par l'utilisateur; Leurs accessoires avec détection des mouvements relatifs en deux dimensions [2D] entre le dispositif de pointage ou une partie agissante dudit dispositif, et un plan ou une surface, p.ex. souris 2D, boules traçantes, crayons ou palets

53.

VEHICLE DAMAGE PICTURE PROCESSING METHOD AND APPARATUS, AND COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020118063
Numéro de publication 2021/218020
Statut Délivré - en vigueur
Date de dépôt 2020-09-27
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liu, Lihong
  • Liu, Yuyu
  • Xiao, Jing

Abrégé

The present application relates to the field of image recognition in artificial intelligence, and provides a vehicle damage picture processing method and apparatus, and a computer device and a storage medium. The method comprises: processing a vehicle damage picture by means of a preset detector, and obtaining the damage detection result of the vehicle damage picture; generating damage intensity graphic data according to the damage detection result; and inputting the damage intensity graphic data into a preset damage intensity classification model for processing, and obtaining the processing result output by the preset damage intensity classification model. According to the present application, the processing capability for the vehicle damage picture can be improved, the robustness of image processing is enhanced, and the accuracy of the image processing result is improved. Moreover, the present application also relates to the blockchain technology, and the present application can be applied to the field of intelligent transportation, thereby promoting the construction of intelligent cities.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

54.

METHOD AND APPARATUS FOR TRAINING NAMED ENTITY RECOGNITION MODEL, AND COMPUTER DEVICE

      
Numéro d'application CN2020118523
Numéro de publication 2021/218024
Statut Délivré - en vigueur
Date de dépôt 2020-09-28
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Chen, Zhenbo
  • Jin, Ge
  • Xu, Liang

Abrégé

The present application relates to the field of artificial intelligence, and provides a method and an apparatus for training a named entity recognition model, and a computer device and a storage medium, comprising: on the basis of a preset training sample, training a BiLSTM-CRF model to obtain three training models; iteratively selecting any two training models randomly from the three training models, and selecting in sequence from an unlabelled data set one item of unlabelled target data to input into the selected two training models for prediction to obtain predicted labels predicted by the two training models; and, if the predicted labels predicted by the two training sets are the same, then adding the predicted labels to the unlabelled target data and updating same in the training sample of the unselected training model in order to train the unselected training model; a semi-supervised method is used to make full use of original data to overcome the defect of an insufficient amount of labelled data; the principle of voting consistency of the three learning models is used to implicitly express the degree of confidence, increasing the model reliability.

Classes IPC  ?

  • G06F 40/295 - Reconnaissance de noms propres
  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion
  • G06N 3/08 - Méthodes d'apprentissage

55.

SDN CONTROLLER-BASED MULTICAST METHOD AND SYSTEM IN HOST OVERLAY NETWORK, DEVICE, AND MEDIUM

      
Numéro d'application CN2020118903
Numéro de publication 2021/218026
Statut Délivré - en vigueur
Date de dépôt 2020-09-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Huang, Bin

Abrégé

The present application relates to the technical field of artificial intelligence. Disclosed are an SDN controller-based multicast method and system in a host overlay network. Multicast group information of an overlay network is controlled in a centralized manner by adopting an SDN controller; a multicast flow rule is generated according to the understanding of the whole network topology by the controller, and issued to a distributed virtual router on a host so as to be executed; when a data plane flow table is delivered, according to obtained interface information, packet information is directly sent to an interface where a multicast group member is located; a flooding problem of a multicast packet on a second layer is well restrained by using an accurate delivery mode, thus avoiding the phenomenon of heavy network burdens.

Classes IPC  ?

  • H04L 12/761 - Routage de diffusion ou de multidiffusion
  • H04L 12/46 - Interconnexion de réseaux
  • H04L 29/12 - Dispositions, appareils, circuits ou systèmes non couverts par un seul des groupes caractérisés par le terminal de données

56.

REPORT GENERATION METHOD AND APPARATUS, AND TERMINAL DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020119297
Numéro de publication 2021/218034
Statut Délivré - en vigueur
Date de dépôt 2020-09-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Lai, Yunhui

Abrégé

Provided are a report generation method and apparatus, and a terminal device and a storage medium, which are suitable for the technical field of computers. The method comprises: acquiring identification information corresponding to a report type of a report to be generated (S101); calling a preset report generation process corresponding to the identification information (S102); acquiring target data corresponding to the identification information according to the preset report generation process, wherein the target data comprises report data and a preset report template (S103); and according to the preset report generation process, filling the preset report template with the report data, so as to obtain a data report corresponding to the report data (S104). All reports share a set of common preset report generation processes to complete the operations of pulling out a common logic and splitting an exception logic, such that code redundancy is reduced. Moreover, it is convenient to add a common logic to a common process without the need to write code for each report logic, such that the commonality and extensibility of the common process are improved.

Classes IPC  ?

57.

DYNAMIC SCENARIO CONFIGURATION-BASED INTERACTIVE PROCESSING METHOD AND APPARATUS, AND COMPUTER DEVICE

      
Numéro d'application CN2020122750
Numéro de publication 2021/218069
Statut Délivré - en vigueur
Date de dépôt 2020-10-22
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Luo, Jinxiong
  • Hu, Hongwei
  • Ma, Jun
  • Wang, Shaojun

Abrégé

Disclosed are a dynamic scenario configuration-based interactive processing method and apparatus, and a computer device. The method comprises: if information to be processed is received from a client, acquiring target character information corresponding to the information to be processed (S110); acquiring interactive scenario information corresponding to the information to be processed (S120); acquiring, in a configuration database, pre-stored configuration information matching the interactive scenario information so that same is used as target configuration information (S130); performing class reflection on a pre-stored general framework according to the target configuration information, so as to construct and obtain a corresponding processing instance (S140); and executing the processing instance to process the target character information to acquire interactive information corresponding to the information to be processed, and feeding back the interactive information to the client to complete interactive processing (S150). The method is based on development assistant technology and relates to the technical field of smart cities, and makes it possible to obtain a corresponding processing instance on the basis of dynamic configuration of interactive scenario information so as to complete interactive processing. The method can be applied to any interactive scenario, and has the characteristics of a wide application range and a higher use flexibility while ensuring the efficiency of interactive processing.

Classes IPC  ?

58.

CALL CONTROL METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020125064
Numéro de publication 2021/218086
Statut Délivré - en vigueur
Date de dépôt 2020-10-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Luo, Jinxiong
  • Hu, Hongwei
  • Ma, Jun
  • Wang, Shaojun

Abrégé

A call control method and apparatus (100), a computer device (500), and a storage medium (503). The method comprises: recording, by means of an incoming call module, the natural language data of a user acquired by an agent terminal (10), and converting the natural language data of the user into corresponding text data (S101); performing word segmentation on the text data to obtain a word segmentation result (S102); using a word in the word segmentation result as an input, and training the word segmentation result of the text data to obtain a word segmentation training result (S103); inputting the word segmentation training result into a pre-trained neural network model for natural language classification, so as to obtain a classification result for the natural language data (S104); and determining a user intention according to the classification result, and determining, by means of the determined user intention, whether to use an outgoing call module to perform a call operation on the user (S105). On the basis of a natural language processing technique in artificial intelligence, the intention of an incoming call of a user can be quickly recognized to help enterprises to improve the customer service quality, thereby improving the customer satisfaction.

Classes IPC  ?

  • G10L 15/22 - Procédures utilisées pendant le processus de reconnaissance de la parole, p.ex. dialogue homme-machine 
  • G10L 15/16 - Classement ou recherche de la parole utilisant des réseaux neuronaux artificiels
  • G10L 15/06 - Création de gabarits de référence; Entraînement des systèmes de reconnaissance de la parole, p.ex. adaptation aux caractéristiques de la voix du locuteur
  • H04M 3/51 - Dispositions centralisées de réponse aux appels demandant l'intervention d'un opérateur
  • G06F 40/289 - Analyse syntagmatique, p.ex. techniques d’états finis ou regroupement
  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion
  • G06N 3/08 - Méthodes d'apprentissage
  • G06N 20/00 - Apprentissage automatique

59.

COMMUNICATION DATA PROCESSING METHOD AND APPARATUS, AND COMPUTER SYSTEM AND STORAGE MEDIUM

      
Numéro d'application CN2020125074
Numéro de publication 2021/218088
Statut Délivré - en vigueur
Date de dépôt 2020-10-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Xi, Yueyang

Abrégé

Disclosed are a communication data processing method and apparatus, and a computer system and a storage medium, which relate to encryption algorithms in the field of blockchains in computer technology, and are applied to a sending end and a receiving end. The method comprises: receiving an enabling signal; a sending end collecting, in real time, timestamp data corresponding to gyroscope data, and synchronously sending the timestamp data to a receiving end; receiving a key generation signal; the sending end and the receiving end respectively acquiring the finally collected gyroscope data as first data and second data, and acquiring initial text to be encrypted; and the sending end and the receiving end respectively processing the first data and the second data by using self-built algorithms to obtain a first key and a second key, wherein the self-built algorithms used by the sending end and the receiving end are the same, the self-built algorithms may be stored in a blockchain node, the sending end and the receiving end respectively perform encryption and decryption, and thus no key needs to be transmitted during data transmission, which solves the problem, in the prior art, of data leakage caused after a key is intercepted in a communication data transmission process.

Classes IPC  ?

60.

SONG SYNTHESIS METHOD, APPARATUS AND DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020131663
Numéro de publication 2021/218138
Statut Délivré - en vigueur
Date de dépôt 2020-11-26
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhu, Qingying
  • Han, Baoqiang

Abrégé

Disclosed is a song synthesis method, comprising: acquiring lyric recitation audio and music score information (101); by means of a preset voice recognition model and lyric pinyin text, performing duration labeling on the lyric recitation audio to obtain a recitation duration (102); analyzing initial acoustic parameters from the lyric recitation audio by means of a preset vocoder (103); according to a preset initials variable-speed dictionary, rhythm information, and beat information, extracting a singing duration from the lyric pinyin text (104); performing variable speed processing on the initial acoustic parameters according to a preset variable-speed algorithm, the recitation duration, and the singing duration (105); performing resonance peak enhancement processing on a spectrum envelope subjected to variable speed processing, so as to obtain an enhanced spectrum envelope (106); performing correction processing on the basis of pitch information, the singing duration, and a fundamental frequency subjected to variable speed processing, so as to obtain a corrected fundamental frequency (107); and performing song synthesis processing on the processed acoustic parameters by means of the preset vocoder (108). The present application further relates to a blockchain, and a synthesized song is stored in the blockchain.

Classes IPC  ?

  • G10L 13/033 - Procédés d'élaboration de parole synthétique; Synthétiseurs de parole Édition de voix, p.ex. manipulation de la voix du synthétiseur
  • G10L 13/08 - Analyse de texte ou génération de paramètres pour la synthèse de la parole à partir de texte, p.ex. conversion graphème-phonème, génération de prosodie ou détermination de l'intonation ou de l'accent tonique
  • G10L 13/10 - Règles de prosodie dérivées du texte; Intonation ou accent tonique
  • G10L 25/18 - Techniques d'analyses de la parole ou de la voix qui ne se limitent pas à un seul des groupes caractérisées par le type de paramètres extraits les paramètres extraits étant l’information spectrale de chaque sous-bande

61.

DATA SYNCHRONIZATION METHOD AND APPARATUS, AND SERVER AND STORAGE MEDIUM

      
Numéro d'application CN2020131952
Numéro de publication 2021/218143
Statut Délivré - en vigueur
Date de dépôt 2020-11-26
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Ding, Yong

Abrégé

Provided are a data synchronization method and apparatus, and a server and a storage medium, which relate to the technical field of data processing. The method comprises: monitoring a modification operation on a data source (S10); when there is a modification operation on the data source, reading a target redo log from a redo log group corresponding to the data source (S11); parsing the target redo log to obtain the type of the target redo log, and acquiring changed data from the target redo log (S12); cleaning the changed data to obtain target data, and storing the target data according to the type (S13); upon the reception of a data synchronization request, extracting engine information from the data synchronization request (S14); determining a storage engine according to the engine information, and establishing a connection with the storage engine to obtain a network link (S15); according to the type and the target data, generating an instruction corresponding to the storage engine (S16); and sending the instruction to the storage engine by means of the network link, such that the storage engine completes data synchronization (S17). The establishment of a network connection with the data source by means of the server reduces the complexity of communication between the data source and the storage engine. The method further relates to blockchain technology. The unified synchronization of data by means of the server can reduce the consumption of CPU and memory resources of a device where the data source is located, and improve the operating efficiency of the device where the data source is located.

Classes IPC  ?

  • G06F 16/22 - Indexation; Structures de données à cet effet; Structures de stockage

62.

HARMONY MATCHING METHOD AND APPARATUS, ELECTRONIC DEVICE, AND COMPUTER READABLE MEDIUM

      
Numéro d'application CN2020134894
Numéro de publication 2021/218158
Statut Délivré - en vigueur
Date de dépôt 2020-12-09
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Jiang, Huijun
  • Huang, Yinxing
  • Jiang, Kaiying
  • Xu, Wei

Abrégé

A harmony matching method and apparatus, an electronic device, and a computer readable medium. The method comprises: dividing a melody segment into multiple target segments on the basis of a meter of the melody segment (S101); respectively calculating, on the basis of note sequences of the respective target segments, matching degrees between multiple preset chords and each target segment, and for each target segment, using a preset chord having the highest matching degree as a target chord of the corresponding target segment (S102); and generating, on the basis of target chords respectively corresponding to the multiple target segments, a harmony segment matching the melody segment (S103). The invention improves the accuracy of matching between harmony segments and melody segments, and improves the smoothness of the overall performance of harmony segments. In addition, the invention further relates to blockchain technology, and allows melody segments to be stored in a blockchain.

Classes IPC  ?

  • G10H 1/38 - Accord
  • G10H 7/00 - Instruments dans lesquels les sons sont synthétisés à partir d'une mémoire de données, p.ex. orgues à calculateur

63.

PAGE DATA GENERATION METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020136016
Numéro de publication 2021/218177
Statut Délivré - en vigueur
Date de dépôt 2020-12-14
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhou, Xueli
  • Zhu, Endong
  • Zhang, Qian
  • Ling, Haiting

Abrégé

A page data generation method, comprising: perform, by means of a preset binning policy, binning processing on data to be binned in a database to obtain binning information and a data partition of the data to be binned (S11); send the binning information to a service terminal to instruct the service terminal to perform a Hash operation on a data sample according to the binning information so as to obtain a binning number (S12); the service terminal performs an engine calculation on sample information corresponding to the binning number and data information of a data partition corresponding to the binning number to obtain an intermediate result (S13); receive the intermediate result sent by the service terminal and perform an engine calculation to obtain a calculation result (S14); the service terminal integrates the calculation result to obtain page data (S15); and a data support terminal performs page displaying according to the page data (S16). The page data is stored in a block chain.

Classes IPC  ?

  • G06F 16/9535 - Adaptation de la recherche basée sur les profils des utilisateurs et la personnalisation
  • G06F 21/60 - Protection de données

64.

AUTOMATIC REPORT GENERATION METHOD, DEVICE, COMPUTER APPARATUS, AND STORAGE MEDIUM

      
Numéro d'application CN2020136079
Numéro de publication 2021/218178
Statut Délivré - en vigueur
Date de dépôt 2020-12-14
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Peng, Yibo
  • Li, Guanjie
  • Luo, Jingjing

Abrégé

An automatic report generation method, a device, a computer apparatus, and a storage medium. The method comprises: generating, in response to a first user operation, a report configuration file comprising a user-defined script address and a task type (S11); generating a user-defined script according to the user-defined script address via Zabbix, and acquiring, by means of the user-defined script, target report data from a target terminal (S12); processing the target report data, and displaying the processed target report data in a report page (S13); upon receiving a task trigger instruction, acquiring, via Puppeteer, the processed target report data from the report page (S14); and parsing the processed target report data, extracting data therefrom, and generating a data report (S15). The invention combines Zabbix-based user-defined scripts and the Puppeteer-based automatic acquisition technique, overcomes restrictions on accessing a target terminal database due to the presence of a bastion host, improves accuracy and efficiency of data acquisition, and enables state monitoring during data acquisition processes, thereby improving error tolerance of acquisition processes.

Classes IPC  ?

65.

ABNORMAL DATA QUERY METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020136080
Numéro de publication 2021/218179
Statut Délivré - en vigueur
Date de dépôt 2020-12-14
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Li, Linge
  • Qi, Mingyuan

Abrégé

An abnormal data query method and apparatus, a computer device, and a storage medium. The method comprises: obtaining a configuration file when abnormal information is detected; screening a target abnormal system according to the abnormal information to obtain a login field in the configuration file, automatically logging in to the target abnormal system according to the login field, and performing query according to a time query field in the configuration file to obtain from the target abnormal system an anomaly time matching the abnormal information; marking the anomaly time as an anomaly start time, and locating, according to the anomaly start time, an anomaly start position from operation data of the target abnormal system; and locating, according to an anomaly end time, an anomaly end position in the operation data of the target abnormal system, and retrieving the time-series operation data located between the anomaly start position and the anomaly end position and marking same as the abnormal data of the target abnormal system. The method can accurately and efficiently locate the abnormal data, thus improving abnormal data processing efficiency.

Classes IPC  ?

  • G06F 16/9535 - Adaptation de la recherche basée sur les profils des utilisateurs et la personnalisation

66.

AI-BASED INTERVIEW CORPUS CLASSIFICATION METHOD AND APPARATUS, COMPUTER DEVICE AND MEDIUM

      
Numéro d'application CN2020093263
Numéro de publication 2021/217772
Statut Délivré - en vigueur
Date de dépôt 2020-05-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Deng, Yue
  • Jin, Ge
  • Xu, Liang

Abrégé

An AI-based interview corpus classification method and apparatus, a computer device, and a readable storage medium. Said method comprises: classifying interviewees by fully considering several corpus features of an interviewees' corpus in an interview scenario related to a prediction result according to a lightweight model obtained by training on the basis of a GPT model, and storing the classification result in a blockchain network node. The interviewees are classified by means of the lightweight model obtained from the GPT model. Because a loss function of each layer of the lightweight model relative to the GPT model is calculated, the accuracy of an output result of the lightweight model and the consistency between the output result of the lightweight model and the output result of the GPT model can be ensured, thereby solving the technical problem of low accuracy of a classification result obtained by using a lightweight network to classify interviewees in the prior art.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06F 16/332 - Formulation de requêtes
  • G06F 40/30 - Analyse sémantique

67.

DATA TRANSMISSION RATE ACQUISITION METHOD, DEVICE, AND COMPUTER APPARATUS

      
Numéro d'application CN2020098792
Numéro de publication 2021/217847
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liu, Mingliang
  • Chen, Youxin
  • Chen, Yuanxu
  • Liu, Weichao
  • Xiao, Jing

Abrégé

The present application pertains to the field of big data, and discloses a data transmission rate acquisition method. The method comprises: acquiring, in real-time, a specified channel quality indicator value in a wireless transmission channel at a current time, acquiring, by means of the specified channel quality indicator value, a transmission rate corresponding to the specified channel quality indicator value, and calculating a basic transmission rate in the wireless transmission channel corresponding to the current time; acquiring a specified time period including the current time and a specified region range covering a mobile terminal connected to the wireless transmission channel, and acquiring a corresponding specified rate conversion factor β according to the specified time period and the specified region range, where 0 ≤ β ≤ 1; and multiplying the basic transmission rate in the wireless transmission channel corresponding to the current time by the specified rate conversion factor to obtain a data transmission rate in the wireless transmission channel corresponding to the current time. The invention achieves accurate data transmission rate acquisition by adding a wireless signal quality sensor at a mobile edge side.

Classes IPC  ?

  • H04W 24/08 - Réalisation de tests en trafic réel
  • H04L 12/26 - Dispositions de surveillance; Dispositions de test
  • H04B 17/309 - Mesure ou estimation des paramètres de qualité d’un canal

68.

DISEASE NAME CODE MATCHING METHOD AND APPARATUS, COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020098937
Numéro de publication 2021/217850
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Jin, Xiaohui
  • Ruan, Xiaowen
  • Xu, Liang

Abrégé

A disease name code matching method and apparatus, a computer device and a storage medium, belonging to the field of artificial intelligence. The method comprises: acquiring a disease name list from an electronic medical record (201); carrying out duplication removal processing on repeated disease names in the disease name list to obtain a duplication-removed disease name list (202); inputting the duplication-removed disease name list into an accurate matching model, and according to a standard disease classification table, carrying out code matching to obtain a first code matching result and candidate code matching disease names (203); inputting the obtained candidate code matching disease names into a fuzzy matching model, and according to the standard disease classification table, carrying out code matching to obtain a second code matching result (204); and according to the first code matching result and the second code matching result, generating a disease name code matching list (205). The method carries out multi-dimensional and multi-mode code matching on disease names, improving the accuracy of disease name code matching; and the disease name list can further be stored in a blockchain to improve data privacy and safety.

Classes IPC  ?

  • G06F 40/216 - Analyse syntaxique utilisant des méthodes statistiques

69.

ABNORMAL ROOT CAUSE POSITIONING METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020098980
Numéro de publication 2021/217855
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Chen, Zhenbo
  • Jin, Ge
  • Xu, Liang

Abrégé

An abnormal root cause positioning method and apparatus, and an electronic device and a computer storage medium. The method comprises: after receiving a root cause positioning instruction which is transmitted by a user and carries alternative instance sets and a specified index, respectively calculating second similarities between alternative instances and the specified index; comprehensively analyzing the second similarities, abnormal time points, and call chain levels, generating an abnormal root cause list on the basis of an analysis result, and feeding back the abnormal root cause list to the user. The efficiency and accuracy of abnormal root cause positioning can be improved.

Classes IPC  ?

  • H04L 12/24 - Dispositions pour la maintenance ou la gestion

70.

FACE IMAGE GENERATION METHOD AND APPARATUS, ELECTRONIC DEVICE, AND READABLE STORAGE MEDIUM

      
Numéro d'application CN2020098982
Numéro de publication 2021/217856
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Tian, Xiao
  • Zhang, Zhanwang
  • Zhou, Chaoyong
  • Liu, Yuyu

Abrégé

A face image generation method and apparatus, an electronic device, and a computer readable storage medium in the technical field of image processing in artificial intelligence. The method comprises: in response to a face image generation request sent out by a client, the request comprising the number m of users whose face images are to be generated and the number n of face images of each user, generating, according to the request, m first random vectors and n second random vectors obeying normal distribution (S1); inputting the m first random vectors into a feature separation region of a face image generation model to obtain a high-order feature vector set (S2); inputting the n second random vectors into the feature separation region of the model to obtain a low-order feature vector set (S3); and matching m first high-order feature vectors in the high-order feature vector set with n second low-order feature vectors in the low-order feature vector set one by one to obtain m*n vector pairs, and inputting each vector pair into an image generation region of the model to obtain m*n face images (S4). The purity and diversity of training set images are improved. The present invention can further be applied to smart security, smart transportation, smart community and other smart city fields, and promotes the construction of smart cities.

Classes IPC  ?

  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales
  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques

71.

SLICE DEFECT DETECTION METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND READABLE STORAGE MEDIUM

      
Numéro d'application CN2020098983
Numéro de publication 2021/217857
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Wang, Jiaping
  • Nan, Yang
  • Li, Fengyi
  • Xie, Chunmei
  • Hou, Xiaoshuai

Abrégé

A slice defect detection method and apparatus, an electronic device, and a readable storage medium, which relate to the technical field of intelligent decision making in artificial intelligence. The method comprises: inputting a slice set to be subjected to detection into a feature extraction branch of a trained slice defect detection model; inputting an obtained feature set into an up-sampling branch of the slice defect detection model, so as to obtain various defect areas and predicted defect categories corresponding to the various defect areas; inputting the feature set into a classification branch of the slice defect detection model, so as to obtain a second defect category distribution table; and acquiring second probability values of the predicted defect categories in the second defect category distribution table, and when the second probability values are greater than a preset threshold value, taking the predicted defect categories as target defect categories. By means of the method, the accuracy of detecting slice defect areas and defect categories is improved. In addition, the method further relates to blockchain technology, and can be applied to the field of smart medical treatment, thereby promoting the construction of a smart city.

Classes IPC  ?

  • G01N 21/88 - Recherche de la présence de criques, de défauts ou de souillures

72.

TARGET IDENTIFICATION METHOD AND APPARATUS BASED ON PICTURE, AND ELECTRONIC DEVICE AND READABLE STORAGE MEDIUM

      
Numéro d'application CN2020098990
Numéro de publication 2021/217858
Statut Délivré - en vigueur
Date de dépôt 2020-06-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Tong, Xinyu
  • Liu, Lihong
  • Liu, Yuyu

Abrégé

A target identification method and apparatus based on a picture, and an electronic device and a computer-readable storage medium. The method comprises: by using a scene segmentation network, executing a convolution operation, an activation operation and a pooling operation on an original picture, so as to obtain a first feature set (S1); in the scene segmentation network, executing an up-sampling operation, the convolution operation and the activation operation on the first feature set, so as to obtain a second feature set, and performing a classification operation on the second feature set according to a pre-constructed classification function, so as to obtain a scene picture set (S2); and inputting the scene picture set into a target identification network to perform target identification, so as obtain a target picture (S3). The method further relates to blockchain technology, and the original picture and the target picture can be stored in a blockchain node. By means of the method, the problem of excessive calculation resources being occupied due to the large amount of calculation required during a target identification process is solved.

Classes IPC  ?

  • G06K 9/34 - Découpage des formes se touchant ou se chevauchant dans la zone image

73.

ORDER IDENTIFIER GENERATION METHOD AND APPARATUS, SERVER, AND STORAGE MEDIUM

      
Numéro d'application CN2020099301
Numéro de publication 2021/217863
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Chen, Ping

Abrégé

An order identifier generation method, comprising: receiving an order determination request sent by a frontend device, the order determination request carrying an order type and a region where an order is located (S11); determining an order number identifier according to the order type, a value weight of an order corresponding to the order type, and the total order amount data of the order, and querying a mechanism identifier corresponding to the region where the order is located from the correspondence between a region where a preset order is located and a preset mechanism identifier (S12); determining an identifier of a server database corresponding to a target database by means of a distributed service framework zookeeper naming service from a plurality of databases comprised by a distributed system (S13); loading a first preset number of first number segments to a remote dictionary server (Redis) cache from a number segment information table of the target database (S14); extracting a target number identifier of the first number segment from the Redis cache (S15); generating an order identifier according to the order number identifier, the mechanism identifier, the identifier of the server database, and the target number identifier (S16); and in response to the order determination request, sending the order identifier to the frontend device (S17). The related data and information can be stored in a blockchain node.

Classes IPC  ?

  • G06F 16/22 - Indexation; Structures de données à cet effet; Structures de stockage

74.

LOAD BALANCING METHOD AND APPARATUS FOR STORAGE CLUSTER, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020099309
Numéro de publication 2021/217864
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Ma, Yuanyuan
  • Gu, Zhongyan
  • Liu, Zirui

Abrégé

A load balancing method for a storage cluster, relating to the field of cloud technology. The method comprises: upon receiving a load balancing instruction, calculating a group-count mean value for basic data of all physical nodes in a storage cluster (S10); determining, according to the group-count mean value and a group count of basic data in each physical node, a migration group count of basic data in each physical node (S20); determining, for each physical node, migration data from the basic data according to the migration group count (S30); acquiring, from all the physical nodes in the storage cluster, a physical node having the smallest group count of basic data to serve as an initial migration-reception node (S40); and if the initial migration-reception node has different basic data, migrating the migration data to the initial migration-reception node so as to perform load balancing (S60). The invention thus improves the performance of storage clusters.

Classes IPC  ?

  • H04L 29/08 - Procédure de commande de la transmission, p.ex. procédure de commande du niveau de la liaison
  • G06F 9/50 - Allocation de ressources, p.ex. de l'unité centrale de traitement [UCT]

75.

GIT-BASED PROJECT VERSION RELEASE METHOD AND APPARATUS, DEVICE, AND MEDIUM

      
Numéro d'application CN2020099527
Numéro de publication 2021/217868
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Geng, Ming

Abrégé

A Git-based project version release method and apparatus, a device, and a medium. The method comprises: upon receiving a project version creation release instruction, obtaining a creation release file package, wherein the creation release file package comprises an initial version number, an initial version file, and an initial deployment operation file associated with each other; storing the initial version file to a Git local repository and pushing same to a Git central repository; sending to first target servers a version initial instruction comprising the initial version number; upon receiving a first pulling instruction of the first target servers, transmitting the initial version file to the first target servers by means of the Git central repository according to a first transmission mode; installing the initial version file to the first target servers according to an initial deployment command file; and upon receiving all installation success instructions, determining that the creation release file package is successfully released. The method implements the automatic configuration of an environment of a target server on the basis of a Git distributed control system, improves efficiency, and reduces costs.

Classes IPC  ?

  • G06F 8/61 - Installation
  • G06F 8/65 - Mises à jour
  • G06F 8/71 - Gestion de versions ; Gestion de configuration
  • G06F 11/14 - Détection ou correction d'erreur dans les données par redondance dans les opérations, p.ex. en utilisant différentes séquences d'opérations aboutissant au même résultat

76.

METHOD AND SYSTEM FOR TRANSFERRING REAL IP ADDRESS OF CLIENT

      
Numéro d'application CN2020099529
Numéro de publication 2021/217869
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Qin, Huawei

Abrégé

The embodiments of the present application disclose a method and a system for transferring a real IP address of a client. Said method comprises: a client sending a communication request to a proxy server, the communication request carrying a client IP packet, the client IP packet comprising a real IP address of the client; the proxy server performing, in response to the communication request, IP packet encapsulation on the client IP packet to obtain a target proxy IP packet corresponding to the client IP packet, the target proxy IP packet comprising the real IP address of the client, and sending the target proxy IP packet to a back-end server on the basis of a communication connection with the back-end server; and the back-end server parsing the target proxy IP packet to obtain the real IP address of the client corresponding to the target proxy IP packet. In addition, the present application also relates to blockchain technology, and a real IP address of a client can be stored in a blockchain node.

Classes IPC  ?

  • H04L 29/12 - Dispositions, appareils, circuits ou systèmes non couverts par un seul des groupes caractérisés par le terminal de données

77.

METHOD AND APPARATUS FOR DEPLOYING MICRO SERVICE CLUSTER, COMPUTER DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020099554
Numéro de publication 2021/217871
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Geng, Ming
  • Jin, Xin

Abrégé

A method and an apparatus for deploying a micro service cluster, a computer device and a storage medium. Said method comprises: upon receipt of a micro service cluster deployment instruction, acquiring a micro service cluster deployment file; storing a target environment configuration package into a local warehouse, making a Docket mirror file from the target environment configuration package by means of a Docker tool, associating the Docker mirror file with an environment configuration package name, and uploading the Docker mirror file to a Docker register; querying, by means of a Helm tool, from a Kubernetes cluster container whether there is an environment cluster name consistent with a target environment cluster name; creating a first micro service cluster consistent with the target environment cluster name; installing, by means of the Helm tool, the target environment configuration package into a first target host, and pushing the Docker mirror file to all of first target nodes; and constructing an environment configuration of the first micro service cluster by means of the Helm tool. The present invention implements automatic configuration of a cluster under a Kubernetes cluster container, reducing manual configuration, and improving the construction efficiency.

Classes IPC  ?

78.

METHOD AND APPARATUS FOR CONFIGURING GATEWAY NODE ON THE BASIS OF VIRTUAL PRIVATE CLOUD, AND MEDIUM

      
Numéro d'application CN2020099560
Numéro de publication 2021/217872
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Qin, Huawei

Abrégé

The present application relates to the field of cloud technology and provides a method for configuring a gateway node on the basis of a virtual private cloud, which is based on a gateway cluster containing a plurality of gateway nodes; a virtual private cloud comprises a plurality of VPC networks, and each gateway node serves a group of VPC networks. The method comprises: determining a target gateway node corresponding to each VPC network, the target gateway node corresponding to each VPC network comprising a main gateway node and at least one backup gateway node; determining the routing priority of each main gateway node and backup gateway node, respectively; and providing a network access service for users in a VPC network on the basis of the main gateway node and at least one backup gateway node corresponding to the VPC network and according to the routing priority; the main network node corresponding to each VPC network synchronizes in real time resource data of a network access service to the corresponding backup gateway node(s). The present application can reduce service interruption caused by gateway node failure and improve user experience.

Classes IPC  ?

  • H04L 12/66 - Dispositions pour la connexion entre des réseaux ayant différents types de systèmes de commutation, p.ex. passerelles
  • H04L 12/707 - Prévention ou récupération du défaut de routage, p.ex. reroutage, redondance de route "virtual router redundancy protocol" [VRRP] ou "hot standby router protocol" [HSRP] par redondance des chemins d’accès
  • H04L 12/703 - Prévention ou récupération du défaut de routage, p.ex. reroutage, redondance de route "virtual router redundancy protocol" [VRRP] ou "hot standby router protocol" [HSRP]

79.

METHOD, APPARATUS AND DEVICE FOR FILTERING OUT BASELINE DRIFT FROM ELECTROCARDIOSIGNAL, AND STORAGE MEDIUM

      
Numéro d'application CN2020099571
Numéro de publication 2021/217873
Statut Délivré - en vigueur
Date de dépôt 2020-06-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Shen, Chaobo
  • Guo, Wei
  • Ruan, Xiaowen
  • Xu, Liang

Abrégé

A method, apparatus and device for filtering out a baseline drift from an electrocardiosignal, and a storage medium. The method for filtering out a baseline drift from an electrocardiosignal comprises: reading an original electrocardiosignal to be processed, and preprocessing the original electrocardiosignal to obtain a first electrocardiosignal sequence; horizontally reversing the first electrocardiosignal sequence, then forwardly inputting the horizontally reversed first electrocardiosignal sequence into a preset IIR filter, and performing interference filtering processing to obtain a second electrocardiosignal sequence; horizontally reversing the second electrocardiosignal sequence, reversely inputting the horizontally reversed second electrocardiosignal sequence into the IIR filter, and performing delay phase correction processing to obtain a third electrocardiosignal sequence; and taking the third electrocardiosignal sequence as an electrocardiosignal subjected to baseline drift interference filtering, and then outputting the electrocardiosignal.

Classes IPC  ?

  • A61B 5/00 - Mesure servant à établir un diagnostic ; Identification des individus

80.

METHOD AND APPARATUS FOR CONTROLLING MECHANICAL ARM ON BASIS OF MONOCULAR VISUAL POSITIONING

      
Numéro d'application CN2020111190
Numéro de publication 2021/217976
Statut Délivré - en vigueur
Date de dépôt 2020-08-26
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Gao, Kaikai
  • Zhou, Chen
  • Zhou, Bao
  • Gong, Lianyin

Abrégé

Provided are a method and apparatus for controlling a mechanical arm on the basis of monocular visual positioning. The method comprises: first, acquiring pixel coordinates of a target point of a target object in a pixel coordinate system; then performing transformation by using a homogeneous transformation matrix of a camera coordinate system with respect to a mechanical arm tail end and a homogeneous transformation matrix of the mechanical arm tail end with respect to a world coordinate system, so as to obtain an equation of a straight line-of-sight; and finally, acquiring, in the world coordinate system, a plane equation of a plane in which the target point is located, and obtaining world coordinates of the target point by means of the equation of the straight line-of-sight and the plane equation. Therefore, compared with the prior art, the embodiments of the present application have no need for two monocular cameras, refine a positioning system, provide simple operations, do not involve the matching of image feature points and the measurement and calculation of an optimal distance, simplify a processing algorithm, and reduce the requirements for a hardware device.

Classes IPC  ?

  • B25J 9/16 - Commandes à programme
  • B25J 19/02 - Dispositifs sensibles
  • G06T 7/80 - Analyse des images capturées pour déterminer les paramètres de caméra intrinsèques ou extrinsèques, c. à d. étalonnage de caméra
  • G06K 9/00 - Méthodes ou dispositions pour la lecture ou la reconnaissance de caractères imprimés ou écrits ou pour la reconnaissance de formes, p.ex. d'empreintes digitales

81.

RAW SPEECH-BASED VOICEPRINT RECOGNITION METHOD, DEVICE, APPARATUS, AND STORAGE MEDIUM

      
Numéro d'application CN2020111439
Numéro de publication 2021/217978
Statut Délivré - en vigueur
Date de dépôt 2020-08-26
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Guo, Yuechao
  • Qiao, Yixuan
  • Tang, Yijun
  • Wang, Jun
  • Gao, Peng
  • Xie, Guotong

Abrégé

The present application pertains to the field of artificial intelligence, and discloses a raw speech-based voiceprint recognition method, capable of reducing information loss and lowering system complexity of raw speech data-based speaker recognition models. The method comprises: acquiring raw speech data, and segmenting the raw speech data according to a preset duration to obtain segmented speech data; performing, by means of a preset convolution filter group, tail-biting convolution processing and discrete Fourier transform processing on the segmented speech data to obtain voiceprint feature data; performing, by means of a preset deep neural network, pooling processing on the voiceprint feature data to obtain a target voiceprint feature; performing embedding vector conversion processing on the target voiceprint feature to obtain a corresponding voiceprint feature vector; and performing computation on the voiceprint feature vector by means of a preset loss function to obtain target voiceprint data, the loss function comprising a cosine similarity matrix loss function and a minimum mean square error matrix loss function. The present application also pertains to blockchain technology, and the voiceprint feature data is stored in a blockchain.

Classes IPC  ?

  • G10L 17/02 - Opérations de prétraitement, p.ex. sélection de segment; Représentation ou modélisation de motifs, p.ex. fondée sur l’analyse linéaire discriminante [LDA] ou les composantes principales; Sélection ou extraction des caractéristiques

82.

VOICEPRINT RECOGNITION METHOD AND APPARATUS, AND DEVICE AND STORAGE MEDIUM

      
Numéro d'application CN2020111464
Numéro de publication 2021/217979
Statut Délivré - en vigueur
Date de dépôt 2020-12-24
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Guo, Yuechao
  • Qiao, Yixuan
  • Tang, Yijun
  • Wang, Jun
  • Gao, Peng
  • Xie, Guotong

Abrégé

The present application relates to the field of artificial intelligence, and discloses a voiceprint recognition method, used for solving the problem of the poor recognition accuracy in voiceprint recognition processes, comprising: acquiring a target voice information set to be recognised, comprising voice information corresponding to at least one subject; using a preset algorithm to extract target feature information from the target voice information set and, on the basis of a first loss function, optimising the target feature information to obtain a first voiceprint recognition result; acquiring target voice channel information of a target voice channel, the target voice channel information comprising channel noise information, and the target voice channel being used for transmitting the target voice information set; extracting a target eigenvector from the channel noise information and, on the basis of the second loss function, optimising the target eigenvector to obtain a second voiceprint recognition result; and merging the first voiceprint recognition result and the second voiceprint recognition result to determine a final voiceprint recognition result. The present application also relates to blockchain technology; the target voice information set is stored in the blockchain.

Classes IPC  ?

  • G10L 17/02 - Opérations de prétraitement, p.ex. sélection de segment; Représentation ou modélisation de motifs, p.ex. fondée sur l’analyse linéaire discriminante [LDA] ou les composantes principales; Sélection ou extraction des caractéristiques
  • G10L 17/06 - Techniques de prise de décision; Stratégies d’alignement de motifs
  • G10L 17/18 - Réseaux neuronaux artificiels; Approches connexionnistes
  • G10L 17/20 - Transformations de motifs ou opérations ayant pout but d’augmenter la robustesse du système, p.ex. contre le bruit du canal ou les différentes conditions de fonctionnement
  • G10L 25/18 - Techniques d'analyses de la parole ou de la voix qui ne se limitent pas à un seul des groupes caractérisées par le type de paramètres extraits les paramètres extraits étant l’information spectrale de chaque sous-bande
  • G10L 25/30 - Techniques d'analyses de la parole ou de la voix qui ne se limitent pas à un seul des groupes caractérisées par la technique d’analyse utilisant des réseaux neuronaux

83.

JAVA CODE PACKING METHOD AND SYSTEM

      
Numéro d'application CN2020111626
Numéro de publication 2021/217980
Statut Délivré - en vigueur
Date de dépôt 2020-08-27
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Yang, Zhe

Abrégé

A java code packing method, comprising: performing, according to a code type and a preset rule, code layering on a source code to be processed to obtain a code to be obfuscated (S102); compiling, by means of a compiling tool, the code to be obfuscated to obtain a compiled code (S104); performing character replacement on the compiled code according to an obfuscation dictionary to generate an obfuscated code (S106); obtaining a pre-generated secret key, and encrypting the obfuscated code by the secret key by means of an encryption rule to obtain an encrypted code (S108); customizing a class loader in a loading container according to the type of the encrypted code, the class loader comprising a decryption rule of the encrypted code (S110); and generating a jar package by using the decryption rule and the encrypted code to complete encryption of a java source program code (S112). The security of the code is improved by means of the approach. In addition, the decryption rule can be stored in a blockchain node.

Classes IPC  ?

  • G06F 21/14 - Protection des logiciels exécutables contre l’analyse de logiciel ou l'ingénierie inverse, p.ex. par masquage

84.

TEXT SUMMARY GENERATION METHOD AND APPARATUS, AND COMPUTER DEVICE AND READABLE STORAGE MEDIUM

      
Numéro d'application CN2020112349
Numéro de publication 2021/217987
Statut Délivré - en vigueur
Date de dépôt 2020-08-31
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zheng, Liying
  • Xu, Liang
  • Ruan, Xiaowen

Abrégé

Disclosed are a text summary generation method and apparatus, and a computer device and a readable storage medium, which relate to artificial intelligence. The method comprises: acquiring text information to be processed (S10); converting the text information into a word vector (S20); inputting the word vector into a pre-trained preset neural network model by means of a beam search algorithm, so as to obtain a candidate summary set of the text information and log likelihood probability values of candidate summaries in the candidate summary set (S30); acquiring target redundancy scores of the candidate summaries in the candidate summary set (S40); acquiring reference scores of the candidate summaries according to the target redundancy scores and log likelihood probability values of the candidate summaries (S50); and selecting, from the candidate summaries, a summary with a reference score greater than a preset reference score so that same is used as a summary corresponding to the text information (S60). In addition, the method further relates to blockchain technology, and the reference scores can be stored in a blockchain. The method can optimize redundant words for an automatically generated summary, thereby improving the readability of an automatically generated text summary.

Classes IPC  ?

  • G06F 16/34 - Navigation; Visualisation à cet effet
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

85.

MOVEMENT CONTROL METHOD AND APPARATUS, STORAGE MEDIUM, AND COMPUTER DEVICE

      
Numéro d'application CN2020112377
Numéro de publication 2021/217988
Statut Délivré - en vigueur
Date de dépôt 2020-08-31
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Gao, Kaikai
  • Zhou, Chen
  • Zhou, Bao
  • Chen, Yuanxu

Abrégé

A movement control method and apparatus, a storage medium, and a computer device, relating to the technical field of artificial intelligence. The invention aims to use actual velocity data of two moving wheels to obtain actual linear velocity data and actual angular velocity data, so as to correct velocity data of the two wheels, thereby reducing velocity errors, and reducing left and right oscillations during the moving process. The control method comprises: receiving a movement control request, and generating a movement control path according to the coordinates of a destination carried in the request (101); processing acquired actual velocity data of two wheels to obtain actual linear velocity data and actual angular velocity data, comparing the actual linear velocity data and the actual angular velocity data respectively with ideal linear velocity data and ideal angular velocity data obtained by means of analyzing the movement control path, and acquiring a comparison result (102); and correcting movement velocity data of the two wheels according to the comparison result, and controlling the two wheels to move to the destination along the movement control path (103). The invention is applicable to movement control.

Classes IPC  ?

  • G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions

86.

METHOD FOR CONSTRUCTING CHINESE-ENGLISH BILINGUAL CORPUS, AND RELATED DEVICE

      
Numéro d'application CN2020117388
Numéro de publication 2021/218012
Statut Délivré - en vigueur
Date de dépôt 2020-09-24
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Deng, Yue
  • Jin, Ge
  • Xu, Liang

Abrégé

A method for constructing a Chinese-English bilingual corpus, and a related device, relating to the technical field of computers, and applied to smart cities, specifically to smart life. The method for constructing the Chinese-English bilingual corpus comprises: obtaining a Chinese entity, an English entity, and a mapping relationship and an intertranslation relationship between the Chinese entity and the English entity, and constructing a bilingual entity word network according to a preset requirement; calculating a single-language representation estimated value and a cross-language entity estimated value of the bilingual entity word network according to the Chinese entity, the English entity, contextual words, a preset hyperlink set, and a preset sentence set; calculating a cross-language sentence estimated value corresponding to an obtained comparable sentence network by using a training sentence; calculating a target estimated value according to the three estimated values; and according to the target estimated value, combining the bilingual entity word network and the comparable sentence network into the Chinese-English bilingual corpus, and storing the Chinese-English bilingual corpus on a blockchain. The accuracy of corpus in the Chinese-English bilingual corpus is improved by using the correlation between the two networks.

Classes IPC  ?

  • G06F 40/58 - Utilisation de traduction automatisée, p.ex. pour recherches multilingues, pour fournir aux dispositifs clients une traduction effectuée par le serveur ou pour la traduction en temps réel

87.

DATA CENTER TRAFFIC MANAGEMENT AND CONTROL METHOD AND SYSTEM

      
Numéro d'application CN2020117887
Numéro de publication 2021/218014
Statut Délivré - en vigueur
Date de dépôt 2020-09-25
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Qin, Huawei

Abrégé

The present application relates to the technical field of big data, and provided is a data center traffic management and control method, which is applied to a data center traffic management and control system, wherein the system comprises a data center and a speed limit cluster which is deployed at the entrance/exit of the data center, with the speed limit cluster comprising a plurality of speed-limiting nodes. The method comprises: a speed-limiting node compiling statistics on traffic of a speed-limited target, so as to obtain a traffic statistical value of the speed-limited target; the speed-limiting node acquiring a preset traffic threshold value corresponding to the speed-limited target; and the speed-limiting node limiting the speed of the speed-limited target when the traffic statistical value of the speed-limited target is greater than the preset traffic threshold value. The data center traffic management and control method realizes, by means of a speed-limiting node, speed limiting management and control of traffic entering and exiting a data center. In addition, further provided is a data center traffic management and control system.

Classes IPC  ?

  • H04L 12/721 - Procédures de routage, p.ex. routage par le chemin le plus court, routage par la source, routage à état de lien ou routage par vecteur de distance
  • H04L 12/741 - Traitement de l'adressage d’en-tête pour le routage, p.ex. table de correspondance
  • H04L 12/801 - Commande de flux ou commande de congestion
  • H04L 12/825 - Commande adaptative, à la source ou aux nœuds intermédiaires, sur réception d'avis de congestion, p.ex. X-on X-off
  • H04L 12/927 - Allocation de ressources en fonction du type de trafic, qualité de service ou priorité
  • H04L 12/939 - Dispositions pour la commutation redondante, p.ex. utilisant des plans de commutation parallèles

88.

PICTURE CROPPING METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020118001
Numéro de publication 2021/218016
Statut Délivré - en vigueur
Date de dépôt 2020-09-27
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Li, Qingpeng

Abrégé

The present application relates to the field of artificial intelligence. Disclosed are a picture cropping method and apparatus, a computer device and a storage medium. Said method comprises: receiving a picture cropping instruction inputted by a user; according to a preset pop-up layer, performing corresponding proportional processing on a first cropping picture; then receiving the size of a cropping frame; determining the actual size of the cropping frame in the first cropping picture; determining the size of a cropped box according to the ratio of the width to the height of the cropping frame; determining the scaling of the first cropping picture; scaling the first cropping picture according to the scaling; obtaining a cropping region of the cropping frame in a target cropping picture; and then according to a left offset and an upper offset of the cropping region, determining the position of the cropping region in the target cropping picture, and hiding an image outside the cropping region. The function of picture cropping is achieved without reading the content of all pixels of the whole picture into a memory and without outputting a new picture again, thereby increasing the picture cropping efficiency.

Classes IPC  ?

  • G06T 7/10 - Découpage; Détection de bords

89.

DATA PROCESSING METHOD AND APPARATUS FOR IMPLEMENTING CROSS-DOMAIN REQUEST AT WEBPAGE END, AND RELATED DEVICE

      
Numéro d'application CN2020118003
Numéro de publication 2021/218018
Statut Délivré - en vigueur
Date de dépôt 2020-09-27
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s) Li, Qingpeng

Abrégé

The present application is applied to the technical field of computers, and disclosed are a data processing method and apparatus for implementing a cross-domain request at a webpage end, a computer device, and a storage medium, for use in improving the expansibility and security of implementing the cross-domain request at the webpage end. The method provided by the present application comprises: receiving, by means of a pre-deployed proxy service, a service request message initiated by a client by means of a webpage, the proxy service being deployed on a server having a same domain name as the webpage, and the proxy service on the server receiving the service request message by means of a pre-created dedicated channel; analyzing an address variable carried by the header of the service request message; obtaining a value of the address variable to obtain a target request address; forwarding, by means of the proxy service, the service request message to a server of a domain name indicated by the target request address; and when a response message returned by the server indicated by the target request address is received, sending, by means of the proxy service, the response message to the client by means of the dedicated channel.

Classes IPC  ?

  • G06F 16/955 - Recherche dans le Web utilisant des identifiants d’information, p.ex. des localisateurs uniformisés de ressources [uniform resource locators - URL]

90.

DATA-BASED BLOOD RELATIONSHIP ANALYSIS METHOD, APPARATUS, AND DEVICE AND COMPUTER-READABLE STORAGE MEDIUM

      
Numéro d'application CN2020118135
Numéro de publication 2021/218021
Statut Délivré - en vigueur
Date de dépôt 2020-09-27
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Huang, Xiangzheng
  • Li, Zhao
  • Wan, Shuwu
  • Li, Jun
  • Zhao, Suqun

Abrégé

The present application relates to the technical field of big data, and discloses a data-based blood relationship analysis method, apparatus, and device and a computer-readable storage medium, used to meet the data-based blood relationship analysis requirements of different types of databases during production practices. The method comprises: acquiring an input table and an output table of structured query language (SQL) statements currently being executed on a big data platform, and a blood relationship between the input table and the output table; converting each of the input table and the output table into an entity object in a system having a pre-configured type, and storing the entity objects in a pre-configured graph database; constructing a blood relationship graph between the entity objects in the graph database according to the blood relationship; receiving a mapping relationship between a service source table and a big data table sent by a data access platform; determining, in the blood relationship graph and according to the mapping relationship, a target entity object node to which an ancestor node is to be added; and adding the corresponding ancestor node to the target entity object node so as to acquire a target blood relationship graph.

Classes IPC  ?

91.

EMOTION DETERMINING METHOD AND APPARATUS FOR MULTIPLE ROUNDS OF QUESTIONS AND ANSWERS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020118464
Numéro de publication 2021/218023
Statut Délivré - en vigueur
Date de dépôt 2020-09-28
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Deng, Yue
  • Jin, Ge
  • Xu, Liang

Abrégé

An emotion determining method for multiple rounds of questions and answers, relating to the field of semantic recognition in artificial intelligence. The method comprises: segmenting a question-and-answer text into several groups of question text segments and several answer text segments; inputting all the question text segments and the answer text segments to a trained model, and respectively encoding each group of the question text segments and answer text segments into matrices, wherein words in the question text segments and the answer text segments are represented by rows or columns in an array; respectively combining all the question text segments with each answer text segment to constitute question-and-answer pairs; updating, according to a correlation between the words in the question text segments and the words in the answer text segments in the question-and-answer pairs, all data represented in the matrices; tiling all the updated matrices, and performing linear transformation to determine a probability of conforming to any emotion type; and taking an emotion having the highest probability as the emotion of an answerer, and storing the emotion of the answerer into a blockchain network.

Classes IPC  ?

  • G06F 16/332 - Formulation de requêtes
  • G06F 40/30 - Analyse sémantique
  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

92.

METHOD AND APPARATUS FOR EXTRACTING TERMINOLOGY IN INTELLIGENT INTERVIEW, DEVICE, AND MEDIUM

      
Numéro d'application CN2020118919
Numéro de publication 2021/218027
Statut Délivré - en vigueur
Date de dépôt 2020-09-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Deng, Yue
  • Jin, Ge
  • Xu, Liang

Abrégé

A method and apparatus for extracting a terminology in an intelligent interview, a computer device, and a storage medium. The method comprises: performing segmentation for each character in an answer statement according to preset N dimensions to obtain N word group sets corresponding to the character, and using all word groups corresponding to the character as candidate word groups; determining weight information of the word group sets according to the word frequency of each word group in a historical interview word bank; assigning weights more conforming to the interview scene to the candidate word groups; determining characterization information of each character by the weight information; and determining, in combination with the characterization information of the characters and by adopting a preset sequence marking model having a simple structure, a named entity comprised in the answer statement, and storing the named entity in a blockchain network as a terminology. There is no need to use a complex named entity extraction model, and thus, the efficiency of determining named entities is improved.

Classes IPC  ?

  • G06F 40/216 - Analyse syntaxique utilisant des méthodes statistiques

93.

ARTIFICIAL INTELLIGENCE-BASED INTERVIEW CONTENT REFINING METHOD, APPARATUS AND DEVICE, AND MEDIUM

      
Numéro d'application CN2020118928
Numéro de publication 2021/218028
Statut Délivré - en vigueur
Date de dépôt 2020-09-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Deng, Yue
  • Jin, Ge
  • Xu, Liang

Abrégé

An artificial intelligence-based interview content refining method, apparatus and device, and a medium. The method comprises: obtaining an interview recording and converting the interview recording into self-introduction text and interview response text (S201); performing text parsing on the self-introduction text, and obtaining basic information of an interviewee (S202); classifying sentences of the interview response text to obtain classified text (S203); using a language extraction model to extract sentences from each type of classified text to obtain extracted sentences, and using a Transformer model to refine the extracted sentences to obtain a refined interview corpus (S204). Thus, the accurate extraction of core content from content of an interview recording having a large amount of data is achieved, and the accuracy of content extraction is improved, which helps to improve the accuracy of intelligent interview evaluation. The basic information of the interviewee and the refined interview corpus are stored in a blockchain and are sent at the same time to a management end for evaluation, which avoids direct semantic recognition that causes evaluation results to fail to meet requirements, which helps to improve the accuracy and efficiency of the intelligent evaluation of interview results.

Classes IPC  ?

  • G06F 40/216 - Analyse syntaxique utilisant des méthodes statistiques
  • G06N 3/04 - Architecture, p.ex. topologie d'interconnexion
  • G06Q 10/00 - Administration; Gestion

94.

ARTIFICIAL INTELLIGENCE-BASED INTERVIEW METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020118949
Numéro de publication 2021/218029
Statut Délivré - en vigueur
Date de dépôt 2020-09-29
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Deng, Yue
  • Jin, Ge
  • Xu, Liang

Abrégé

An artificial intelligence-based interview method and apparatus, a computer device, and a storage medium. The method comprises: obtaining a response text; according to a preset condition, performing semantic recognition and classification on the response text so as to obtain first data, second data, and third data; and inputting the first data, the second data, and the third data to a preset intelligent interview model for training, so as to obtain a response text classification result, a named entity, and an open interview question. Therefore, the present application simultaneously implements a plurality of interview data processing tasks by using one model, ensures the uniformity and completeness of the data, and improves the efficiency of an intelligent interview and the accuracy of an interview result; moreover, the present application determines an instant response score according to the response text classification result, determines an authenticity score of a response according to the named entity, continues to guide the intelligent interview according to the open interview question, and determines an interview result according to the finally obtained instant response score and the authenticity score, thereby achieving intelligent interview question generation and interview guide, and improving the interview efficiency.

Classes IPC  ?

95.

DICTIONARY DATA OPERATION METHOD AND APPARATUS, READABLE STORAGE MEDIUM, AND TERMINAL DEVICE

      
Numéro d'application CN2020119296
Numéro de publication 2021/218033
Statut Délivré - en vigueur
Date de dépôt 2020-09-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Huang, Xinguan
  • Lai, Yunhui
  • Sun, Cheng
  • Zhang, Hao
  • Long, Juegang
  • Zhao, Xianjin
  • Mei, Han

Abrégé

A dictionary data operation method and apparatus, a non-volatile computer readable storage medium, and a terminal device. The method comprises: searching for, in a target program to be executed, a preset dictionary data operation identifier (S101), the dictionary data operation identifier being used for indicating the type of a dictionary data operation; determining a dictionary data operation instruction in said target program (S102), the dictionary data operation instruction being an instruction corresponding to the found dictionary data operation identifier; and extracting a dictionary data parameter in the dictionary data operation instruction, and executing the dictionary data operation instruction according to the dictionary data parameter and the type of the dictionary data operation (S103). According to the method, the addition of a large number of dictionary data operation codes irrelevant to a service is avoided, and the coupling degree of the codes is reduced, thereby reducing the subsequent code maintenance costs.

Classes IPC  ?

  • G06F 16/22 - Indexation; Structures de données à cet effet; Structures de stockage

96.

SMART ROBOT DEPLOYMENT METHOD, APPARATUS, DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020122325
Numéro de publication 2021/218061
Statut Délivré - en vigueur
Date de dépôt 2020-10-21
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Liang, Bizhi
  • Hu, Hongwei
  • Ma, Jun
  • Wang, Shaojun

Abrégé

the present application relates to artificial intelligence, and disclosed within are a smart robot deployment method, an apparatus, a device, and a storage medium. The method comprises: performing intent recognition and slot value filling on voice information to be processed by means of a robot system, and obtaining intent information; performing analysis on the intent information by means of an intent condition table, obtaining target intent information, and acquiring an intent type of the target intent information; traversing a robot structure tree in a preconfigured robot registration center, obtaining a robot and a service address of the robot, and deploying the robot according to the service address; performing buffering on interaction information, and obtaining robot interaction information, where said interaction information may also be stored on a blockchain network; acquiring a target intent type for target voice information, and triggering an intent jump instruction according to the target intent type and a preconfigured jump device; deploying a target robot corresponding to the target intent type and according to the intent jump instruction, and acquiring robot interaction information corresponding to the target intent type.

Classes IPC  ?

97.

METHOD AND APPARATUS FOR ASSIGNING PARAMETER WITH VALUE ON THE BASIS OF NAME MAPPING, AND COMPUTER DEVICE

      
Numéro d'application CN2020122452
Numéro de publication 2021/218063
Statut Délivré - en vigueur
Date de dépôt 2020-10-21
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Luo, Jinxiong
  • Hu, Hongwei
  • Ma, Jun
  • Wang, Shaojun

Abrégé

A method and an apparatus for assigning a parameter with a value on the basis of name mapping, and a computer device. Said method comprises: traversing a program to be traversed inputted by a user to obtain an interface parameter information table; establishing a mapping according to parameter names in the parameter information table to obtain first name mapping information and second name mapping information; storing the first name mapping information, the second name mapping information and an unmapped parameter information table in a parameter database; acquiring, from the parameter database, a target parameter name corresponding to a parameter processing request; and assigning, according to the target parameter name, a corresponding target parameter in said program with a value. Said method is based on a system performance optimization technique, and establishes name mapping information according to parameter names in a parameter information table, so that a storage space occupied by parameter information in a system memory can be reduced. Establishing name mapping information can improve the efficiency and accuracy of data transfer across interfaces, i.e. improving the efficiency of assigning parameters of an interface with values.

Classes IPC  ?

  • G06F 16/31 - Indexation; Structures de données à cet effet; Structures de stockage

98.

INTENTION RECOGNITION METHOD AND APPARATUS BASED ON ARTIFICIAL INTELLIGENCE, AND COMPUTER DEVICE

      
Numéro d'application CN2020125065
Numéro de publication 2021/218087
Statut Délivré - en vigueur
Date de dépôt 2020-10-30
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Fan, Canhua
  • Hu, Hongwei
  • Ma, Jun
  • Wang, Shaojun

Abrégé

An intention recognition method and apparatus based on artificial intelligence, and a computer device, relating to an artificial intelligence technology. The method comprises: obtaining input information of a user (S110); processing the input information to obtain words in the input information (S120); obtaining, according to the words in the input information and a preset domain knowledge graph, a knowledge graph matching the input information, wherein the domain knowledge graph comprises at least one knowledge graph (S130); and determining the number of the knowledge graphs obtained from the domain knowledge graph to obtain the knowledge graph conforming to the input information, thereby obtaining an intention of the user. The method solves, on the basis of a knowledge graph technology, the problems in the prior art of low accuracy and low expansibility during intention recognition, and improves the efficiency in the technical field of intention recognition.

Classes IPC  ?

  • G06F 40/289 - Analyse syntagmatique, p.ex. techniques d’états finis ou regroupement

99.

VOICE-BASED USER GENDER AND AGE RECOGNITION METHOD AND APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

      
Numéro d'application CN2020131612
Numéro de publication 2021/218136
Statut Délivré - en vigueur
Date de dépôt 2020-11-26
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhao, Jing
  • Wang, Jianzong

Abrégé

The present application relates to the technical field of voice classification in artificial intelligence, and provides a voice-based user gender and age recognition method and apparatus, a computer device, and a storage medium. The method comprises: preprocessing received current user voice data transmitted by a user side to obtain preprocessed voice data; performing feature extraction in terms of short-time average amplitude, Mel-frequency cepstral coefficient, and Mel-frequency cepstral coefficient first-order difference on each frame of voice data to obtain corresponding mixed parameter features, so as to form a mixed parameter feature time sequence; inputting the mixed parameter feature time sequence into a Gaussian mixture model to obtain a corresponding current user classification result; and calling a voice response strategy, obtaining corresponding current voice response data, and sending the current voice response data to the user side. Accurate gender and age recognition based on user voice is implemented.

Classes IPC  ?

  • H04M 3/51 - Dispositions centralisées de réponse aux appels demandant l'intervention d'un opérateur
  • G10L 15/02 - Extraction de caractéristiques pour la reconnaissance de la parole; Sélection d'unités de reconnaissance 
  • G10L 15/06 - Création de gabarits de référence; Entraînement des systèmes de reconnaissance de la parole, p.ex. adaptation aux caractéristiques de la voix du locuteur
  • G10L 15/22 - Procédures utilisées pendant le processus de reconnaissance de la parole, p.ex. dialogue homme-machine 
  • G10L 15/24 - Reconnaissance de la parole utilisant des caractéristiques non acoustiques

100.

DEFORMABLE CONVOLUTION-BASED IMAGE RECOGNITION METHOD AND APPARATUS, AND COMPUTER DEVICE

      
Numéro d'application CN2020131756
Numéro de publication 2021/218140
Statut Délivré - en vigueur
Date de dépôt 2020-11-26
Date de publication 2021-11-04
Propriétaire PING AN TECHNOLOGY (SHENZHEN) CO., LTD. (Chine)
Inventeur(s)
  • Zhao, Xiaohong
  • Liu, Lihong
  • Liu, Yuyu

Abrégé

A deformable convolution-based image recognition method and apparatus, a computer device, and a storage medium. The method comprises: adjusting a first bias domain by using a training image; performing deformable convolution on the training image according to the first bias domain and a convolution layer to obtain a training feature image; adjusting a second bias domain according to the training feature image and the training image to obtain a trained deformable convolutional neural network; and recognizing an image to be recognized according to the deformable convolutional neural network after multiple trainings to obtain a corresponding recognition result, and feeding same back to a client. The present method is based on neural network technology, wherein by means of the deformable convolutional neural network that comprises the first bias domain and the second bias domain, the image is recognized by using the deformable convolution, and graphics to be recognized that have discontinuous pixels in the image can be recognized, thereby greatly improving the accuracy of recognizing irregular graphics comprised in the image.

Classes IPC  ?

  • G06K 9/62 - Méthodes ou dispositions pour la reconnaissance utilisant des moyens électroniques
  1     2     3     ...     50        Prochaine page