BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xu, Ran
Abstract
Provided in the present application are a three-dimensional road network construction method and apparatus, and an electronic device and a storage medium, the method being executed by the electronic device. The method comprises: acquiring a two-dimensional road network, and in all nodes of a path, setting a corresponding relative elevation for a node which overlaps with the current path and/or other paths; reducing the relative elevation of a target node having the largest relative elevation in the path, and increasing the relative elevations of the other nodes in the path; determining a center line of the path, and according to the relative elevations of the nodes of the path, determining the relative elevations of nodes that constitute the center line, and setting the width of the path; and according to the width of the path and the relative elevations of the nodes of the center line, constructing a three-dimensional road which corresponds to the path, so as to obtain a three-dimensional road network.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Meng, Fanbo
Liu, Jinsuo
Zhu, Weiji
Zhang, Yongzhe
Feng, Tian
Abstract
A video processing method and apparatus, and a medium and a program product. The method specifically comprises: acquiring a first video clip, wherein the first video clip corresponds to template text in first text of a video to be generated, the first video clip comprises a video sub-clip of which voice is paused, and the position of the video sub-clip corresponds to a demarcation position between the template text and variable text to be processed in the first text (101); generating a second video clip corresponding to the variable text to be processed (102); and splicing the first video clip and the second video clip, so as to obtain a video corresponding to the first text (103). By means of the embodiments of the present application, the processing efficiency of a video can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Fan, Ziwei
Zhan, Jiqing
Yu, Jian
Wang, Yanfeng
Zhu, Yun
Zhao, Ang
Abstract
The present application discloses a data processing method applied in fields such as artificial intelligence and natural language processing. Inputting into a pre-trained model and an initial service model on the basis of acquired service data, to obtain a first fully connected layer output of the pre-trained model and a second fully connected layer output of the initial service model. Because the pre-trained model and the initial service model construct modeling units using different granularities, a first modeling unit of the pre-trained model is matched with a second modeling unit of the initial service model, fully connected layer output of a second modeling unit and fully connected layer output of a matching target first modeling unit have the same sequence length, and knowledge distillation is carried out on this basis, a target service model can be obtained by means of the initial service model with the assistance of the pre-trained model, and knowledge distillation is carried out on a model comprising modeling units of different granularities, thereby expanding an application range for knowledge distillation, and effectively improving service performance of the service model.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Duan, Wenjun
Bi, Jingrong
Li, Mei
Meng, Fanbo
Peng, Yun
Liu, Kai
Abstract
A data processing method and apparatus, and a computer device and a storage medium. The method comprises: acquiring sign language action data to be processed (101); on the basis of a preconstructed sign language annotation system, performing element analysis on the sign language action data, so as to determine a sign language annotation sequence, which corresponds to the sign language action data, wherein the sign language annotation sequence contains annotation information of each basic sign language element, which corresponds to the sign language action data (102); and according to the sign language annotation sequence, performing operation processing on the sign language action data (103).
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Tian, Kai
Chen, Wei
Su, Xuefeng
Abstract
An action processing method and apparatus for a virtual object, and a storage medium. The method specifically comprises: receiving an action instruction, wherein the action instruction may comprise: an action identifier and time-related information (101); determining an action video frame sequence corresponding to the action identifier (102); according to a preset state image of a virtual object at a target time, determining, from the action video frame sequence, an action state image corresponding to the preset state image, wherein the target time can be determined according to the time-related information (103); according to the preset state image and the action state image, generating a connecting video frame sequence, wherein the connecting video frame sequence is used for connecting the preset state image and the action video frame sequence (104); and splicing the connecting video frame sequence and the action video frame sequence to obtain an action video (105). By means of the method, the action processing efficiency for a virtual object is improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Fang, Peng
Liu, Kai
Chen, Wei
Abstract
A speech synthesis method, an apparatus, a device, and a storage medium. The method comprises: acquiring a text to be synthesized (101); generating a hidden layer feature and a prosodic feature of the text to be synthesized, and predicting the pronunciation duration of a character in the text to be synthesized (102); on the basis of the hidden layer feature, the prosodic feature, and the pronunciation duration, generating an acoustic feature corresponding to the text to be synthesized (103); and according to the acoustic feature, generating text speech corresponding to the text to be synthesized (104). The described method is helpful to reduce the difficulty of speech synthesis.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Cui, Guohui
Abstract
Embodiments of the present application provide a speech processing method and apparatus, and an apparatus for speech processing, applied to a terminal device. The terminal device is provided with at least two microphones. The method comprises: performing summation processing on signals received by the at least two microphones to obtain a first signal, and performing subtraction processing on the signals received by the at least two microphones to obtain a second signal; performing blind separation processing on the first signal and the second signal to obtain a speech signal and a noise signal; and performing adaptive noise cancellation processing on the speech signal on the basis of the noise signal to obtain a target speech signal. According to the embodiments of the present application, a speech denoising effect can be optimized, so that the accuracy of speech recognition of the terminal device is improved under the condition that the environment is complex and changeable and noise or interference is large.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Jie
Wang, Yingjun
Qin, Bo
Xin, Xiaozhe
Abstract
Disclosed in the present application is a formula recognition method. In one example, the method may be executed by a first device. The first device may obtain an image to be recognized that comprises a formula, and then process said image to obtain a global image feature and a local image feature. Furthermore, the global image feature and the local image feature may be processed to obtain the formula comprised in said image. Therefore, in embodiments of the present application, during recognition of the formula in said image, the local image feature of said image is combined. Hence, even if characters comprised in the formula in said image are small, the formula in said image can also be accurately recognized using the solutions of the embodiments of the present application. Therefore, the formula in said image can be accurately recognized using this solution.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Han, Qin
Yu, Tianzhao
Wang, Yu
Sheng, Yue
Zhuo, Xingzhong
Abstract
Embodiments of the present application provide a request processing method and apparatus, and a medium. The method specifically comprises: presenting a candidate item, and timing the presentation time; and sending a request when the timing value is within a time range, wherein the request is used for requesting a cloud candidate; the time range is obtained according to historical behavior data corresponding to the candidate item; the historical behavior data comprises: on-screen behavior data when presenting the candidate item, and/or trigger behavior data for a preset control when presenting the candidate item; and the preset control is used for triggering presentation of more candidate items. The embodiments of the present application can improve the speed of presenting a cloud candidate while reducing the burden on a server.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Guo, Baokui
Kang, Qi
Abstract
Embodiments of the present application provide a poem generation method and apparatus, and a medium. The method specifically comprises: receiving generated information; and according to an auto-regressive language model, determining at least one candidate poem corresponding to the generated information, wherein the language model is trained and obtained by means of a poem corpus, and is used to predict unknown information of a poem in the unit of words according to known information of the poem. The language model comprises a plurality of processing layers that are sequentially connected. The processing layers comprise a self-attention module and a neural network module, wherein the self-attention module is used to determine information of attention from known words in a sentence of a poem to words in a word list, so as to predict unknown words in the sentence of the poem according to the information of attention. By means of the embodiments of the present application, a candidate poem which is in compliance with rules of poems can be generated, and the coherence of the generated candidate poem can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
He, Junqing
Abstract
Embodiments of the present application provide an inquiry information processing method and apparatus, and a medium. The method specifically comprises: determining disease features of a user according to at least one user input; performing disease prediction processing on the disease features of the user, so as to obtain corresponding candidate diseases; acquiring corresponding question entities from a knowledge graph according to disease feature entities corresponding to the disease features of the candidate diseases, wherein the question entities are used for representing questions related to the disease feature entities; and generating target questions according to questions corresponding to the question entities, wherein the target questions are used for performing an inquiry on the user. By means of the embodiments of the present application, the accuracy of target questions for an inquiry can be improved, such that the inquiry efficiency is improved.
G16H 80/00 - ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring
12.
METHOD AND APPARATUS FOR SETTING APPLICATION PROGRAM
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhang, Yijun
Abstract
A method and an application for setting an application program, specifically: in order to enable a user to clearly view target content in a page corresponding to an application program, acquiring the background colour of the target content; and then determining the page content colour of the application program on the basis of the acquired background colour, wherein the contrast between the determined page content colour and the background colour is greater than or equal to a first threshold. In other words, the contrast between the page content colour and the background colour is greater. After determining the page content colour of the application program, implementing colour setting of the target content according to the page content colour, such that the colour of the target content is the determined page content colour.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xin, Xiaozhe
Qin, Bo
Zhao, Zhiyong
Wang, Yingjun
Wang, Jie
Su, Xuefeng
Chen, Wei
Abstract
A handwriting recognition method, comprising: acquiring handwritten original trajectory data in real time (S101); compressing the handwritten original trajectory data to obtain compressed handwritten trajectory data (S102); and inputting the compressed handwritten trajectory data into a compressed handwriting recognition model for recognition to obtain a text recognition result corresponding to the handwritten original trajectory data, the handwriting recognition model being obtained by training handwritten trajectory data of each piece of training data in a training data set, and the compressed handwriting recognition model being obtained by performing model compression on the handwriting recognition model (S103). The method can effectively improve accuracy of handwriting recognition.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
He, Fulin
Abstract
The embodiments of the present application disclose an input method and device, and a device for input. The embodiment of said method comprises: detecting, in real time, a contact position during handwriting input of a user; on the basis of the contact position and information related to the user, determining a display point position, the display point position being offset from the contact position; and on the basis of the display point position, presenting a contact movement trace. The embodiment improves the efficiency and accuracy of handwriting input.
G06F 3/0488 - Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser using a touch-screen or digitiser, e.g. input of commands through traced gestures
15.
SEARCH METHOD AND APPARATUS, AND ELECTRONIC DEVICE
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wei, Tao
Rao, Xudong
Abstract
A search method and apparatus, and an electronic device. The method comprises: acquiring network construction information corresponding to a target task, wherein the network construction information comprises: search space information, sample data and a search index (102); constructing a super-network on the basis of the search space information, and training the super-network on the basis of the sample data, wherein the super-network comprises a plurality of sub-networks (104); and searching the trained super-network for a sub-network on the basis of the search index, so as to obtain a target network for executing the target task (106). Compared with the fact that it is necessary to manually design a network in the prior art, the present application can be adapted to tasks, such that a network can be quickly constructed.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhao, Hang
Abstract
A data processing method and device. The method comprises: determining a search keyword according to a search statement inputted by a user (S101), and determining a first graph entity library corresponding to a category to which the search keyword belongs (S102); according to an association relationship among multiple categories of graph entity libraries, determining the category of a graph entity library associated with the first graph entity library, and constructing a first set, the first set comprising at least one graph entity library having an association relationship with the first graph entity library, and the association relationship among the graph entity libraries being constructed according to search requirements of the user (S103); and after the first set is constructed, using the search keyword as a center to display the graph entity libraries comprised in the first set (S104). The first set can reflect the search requirements of the search statement and guide the user to further trigger a search behavior on the basis of the search requirements of the user; and the method allows for providing a search result meeting the user requirements to the user.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Congtao
Yang, Jiajun
Wei, Yuanming
Chen, Wei
Wei, Tao
Wu, Jun
Gong, Li
Zhu, Weiji
Abstract
An image processing method, comprising: when an image recognition apparatus collects images in real time, storing the collected real-time images into an image splicing module by means of a first thread (S101); performing real-time splicing on the images in the image splicing module by means of a second thread to obtain an intermediate spliced image (S102); performing character recognition on the intermediate spliced image by means of a lightweight optical character recognition model to obtain an intermediate recognition result, and displaying same (S103); and when it is detected that a state of the image recognition apparatus is changed, obtaining an image set consisting of all images in the image splicing module, performing character recognition on the image set by means of a universal optical character recognition model to obtain a final recognition result, and displaying same (S104). According to the image processing method, the image recognition apparatus, an electronic device, and a medium, recognition real-time performance of character recognition for the intermediate spliced image can be improved, and a response duration of on-screen display of a first character is shortened.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Congtao
Yang, Jiajun
Wei, Yuanming
Chen, Wei
Wei, Tao
Wu, Jun
Gong, Li
Zhu, Weiji
Abstract
A picture recognition method and apparatus, an electronic device, and a medium. The method comprises: in the process of performing text recognition on an acquired image by means of an optical character recognition model, obtaining the current recognized character string and a hash value thereof, and storing the current character string and the hash value thereof into a first preset tree-like structure and a second preset tree-like structure; predicting a new probability value of the current character string at a next moment; obtaining an extended character string set; obtaining, according to the hash value stored in the second preset tree-like structure, N character strings having highest probability values, retaining the N character strings, and using the N character strings as the current character strings; and repeating the foregoing steps until the recognition of all acquired images is completed, and obtaining a character string having the highest probability value as a final recognition result. The calculation amount required for storing an intermediate character string in the process of performing text recognition on an image can be effectively reduced, and the efficiency of storing the intermediate character string can be effectively improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Xin
Abstract
An entry recommendation method and apparatus, and an apparatus for recommending entries. The method comprises: recalling a plurality of word lists on the basis of a multi-path recall policy, wherein entries in each word list have original scores; for each word list, respectively performing normalization processing on the original scores so as to obtain target scores of the entries in each word list, and respectively determining the weights of the entries in each word list; and on the basis of the target scores and the weights, selecting a target entry from the plurality of word lists, and pushing the target entry to a target client. By the means of the method, entries recommended to a user can satisfy input requirements of the user to a greater extent, and the accuracy of the entries recommended to the user is improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Liu, Yun
Abstract
Disclosed are a speech processing method and apparatus, and an apparatus for processing speech. An embodiment of the method comprises: acquiring a first spectrum of noisy speech in a complex number field; performing sub-band decomposition on the first spectrum to obtain a first sub-band spectrum in the complex number field; processing the first sub-band spectrum on the basis of a pre-trained noise reduction model, so as to obtain a second sub-band spectrum, in the complex number field, of target speech in the noisy speech; performing sub-band restoration on the second sub-band spectrum to obtain a second spectrum in the complex number field; and synthesizing the target speech on the basis of the second spectrum. By means of the embodiment, the problem of high-frequency and low-frequency information being imbalanced is effectively solved, and the clarity of speech after noise reduction is thus improved.
G10L 25/30 - Speech or voice analysis techniques not restricted to a single one of groups characterised by the analysis technique using neural networks
21.
INPUT METHOD AND APPARATUS, AND APPARATUS FOR INPUTTING
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Yao, Bohuai
Cui, Xin
Abstract
Disclosed in the embodiments of the present application are an input method and apparatus, and an apparatus for inputting. The embodiment of the method comprise: acquiring a first statement inputted by a user; inputting the first statement into a pre-trained rewriting model to obtain a second statement which has the same semantic but a different style with the first statement; and displaying the second statement. The embodiment of the present application can improve the generalization of a statement rewriting function and the fluency of a rewritten statement.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Fang, Fei
Abstract
A word processing method and apparatus, and an apparatus for word processing. Said method specifically comprises: splitting a word inputted by a user, the obtained splitting result comprising at least one sub-word and at least one affix (201); and using a tree structure to display the word and the splitting result hierarchically (202). Said method can improve word learning efficiency, and can reduce word learning difficulty.
G06F 40/284 - Lexical analysis, e.g. tokenisation or collocates
G06F 3/0481 - Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Lin, Jiansu
Liu, Fang
Meng, Yuqing
Huang, Xiaofeng
Chen, Weipeng
Gong, Zhaohan
Zong, Ming
Chen, Liwei
Wang, Dandan
Ning, Yanan
Abstract
A data processing method and device. The method comprises: obtaining a search keyword of a user (S101), determining a plurality of target keywords related to the search keyword (S102), and after determining a plurality of keywords, further displaying the search keyword and the plurality of target keywords, wherein when displaying the search keyword and the plurality of target keywords, the plurality of target keywords are dispersedly distributed on the periphery of a circle taking the search keyword as the center (S103).
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Ying
Li, Jiantao
Zhang, Dan
Liu, Bao
Zhang, Shuo
Yang, Tianfu
Liang, Xiao
Rong, Hejiang
Li, Pengchong
Abstract
A translation method and apparatus, and a headset, with the method being applied to the headset. The method comprises: a headset acquiring source speech data (102); and the headset translating the source speech data to generate target speech data (104). Therefore, a user only needs to use a headset device to be able to realize translation, without using a specialized translation device.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhao, Nan
Cui, Wenhua
Abstract
Provided in the embodiments of the present application are a process execution method, a device, and a readable medium. The method comprises: detecting a work state of a pair of headphones, identifying a movement state of a user via an accelerometer built in the headphones, determining a target process associated to the movement state in the work state, and executing the target process, thus implementing the triggering of various target processes by the headphones on the basis of the movement state of the user, obviating the need for the user to trigger via a hand movement or speaking, and increasing the degree of convenience of triggering various target processes; moreover, target processes corresponding to same movement state of the user may be different in different work states of the headphones, implemented on the headphones is the triggering of more target processes, thus solving the problem in which it is difficult for the user to trigger a target process by means of a hand movement or speaking in some scenarios.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Chen, Wei
Fan, Bo
Meng, Fanbo
Abstract
A method for driving a virtual human in real time, comprising: obtaining data to be processed used for driving a virtual human, said data comprising at least one of text data and voice data (S201); using an end-to-end model to process said data, so as to determine a gesture feature sequence corresponding to said data (S202); and inputting the gesture feature sequence into a trained muscle model to drive the virtual human by means of the muscle model (S203). Since the end-to-end model can obtain the gesture feature sequence in a shorter period of time, and the gesture feature sequence is input into the muscle model to directly drive the virtual human, the amount of computation and the amount of data transmission are greatly reduced, and the computational efficiency is further improved, so that the real time nature of virtual human driving is greatly improved, thereby achieving real time driving of the virtual human to perform sign language output.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Song, Yingshuang
Xie, Shuo
Zhou, Zheng
Xie, Zhuo
Song, Jinchang
Abstract
A scanning and shadow reading processing method and apparatus, a dictionary pen, and a machine-readable medium. The method comprises: in response to a scanning operation of a dictionary pen on text, obtaining a query result of the text and displaying same on a display interface of the dictionary pen (201); in response to a shadow reading speech input operation of a user on the display interface of the query result, obtaining shadow reading speech of the user (202); and obtaining a shadow reading evaluation result of the shadow reading speech on the basis of the shadow reading speech and standard reading speech corresponding to the shadow reading speech, and displaying the shadow reading evaluation result on the display interface of the dictionary pen (203). In this way, by means of a dictionary pen, shadow reading speech of a user can be acquired and evaluated, thereby better assisting the user in learning a query result of text.
G10L 25/51 - Speech or voice analysis techniques not restricted to a single one of groups specially adapted for particular use for comparison or discrimination
G06K 9/32 - Aligning or centering of the image pick-up or image-field
G06F 3/0487 - Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Fan, Bo
Abstract
A method and apparatus for driving a digital person, and an electronic device. The method comprises: acquiring a target action corresponding to target text (S20); obtaining, when driving a digital person to output a voice on the basis of the target text, a reference action needing to be performed by the digital person before performing the target action (S22); modifying a target action parameter of the target action according to a reference action parameter of the reference action (S24); and in the process of driving the digital person to output the voice on the basis of the target text, driving the digital person to perform the target action according to the modified target action parameter (S26). In the technical solution, a corresponding action is obtained on the basis of text, and an action parameter of a target action corresponding to the text is modified according to a reference action of a digital person, so that the process of the digital person switching from the reference action to the target action is natural and delicate, the technical problem in the prior art that the action of the digital person changes suddenly is solved, and the delicacy of change of the action of the digital person is improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Fan, Bo
Chen, Wei
Chen, Xi
Meng, Fanbo
Liu, Kai
Zhang, Kening
Duan, Wenjun
Abstract
Disclosed is a method for driving a virtual human in real time. The method comprises: acquiring data to be processed that is used for driving a virtual human, wherein said data comprises at least one of text data and speech data (S201); processing said data by means of an end-to-end model, so as to determine an acoustic feature sequence, a facial feature sequence and a limb feature sequence corresponding to said data (S202); and inputting the acoustic feature sequence, the facial feature sequence and the limb feature sequence into a trained muscle model, and driving the virtual human by means of the muscle model (S203). In this way, the computing efficiency is improved, and the real-time performance of driving a virtual human is greatly improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Fan, Bo
Xu, Zhen
Chen, Xi
Abstract
The present disclosure discloses a virtual anchor face swapping method and apparatus, an electronic device, and a storage medium. The method comprises: using an end-to-end sequence learning model to process a historical video material of a target virtual anchor so as to acquire face feature parameters of the target virtual anchor; using a candidate virtual face to perform face swapping on the target virtual anchor so as to acquire an image material corresponding to the face-swapped virtual anchor; using the end-to-end sequence learning model to process the face feature parameters and the image material of the target virtual anchor so as to acquire face feature parameters of the face-swapped virtual anchor; and merging the candidate virtual face into the historical video material according to the face feature parameters of the face-swapped virtual anchor so as to acquire a video material resulting from the face swapping. The invention improves the utilization rate of the historical video material, and ensures that an image of the face-swapped virtual anchor is realistic and natural, and that facial features are natural and coordinated.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Dongliang
Liu, Kai
Zhou, Ming
Chen, Wei
Abstract
Disclosed are a speech processing method and apparatus, and an electronic device (800). The method comprises: acquiring text information to be converted (102), and determining a source language corresponding to the text information and a target user for whom conversion is to be carried out (104); and according to the text information and a target conversion model corresponding to said target user, converting the text information into target speech data, which is pronounced in the source language by said target user, wherein the target conversion model performs adaptive training on a trained universal conversion model according to monolingual speech data pronounced by said target user, the universal conversion model is trained according to speech data in N languages, the source language is one of the N languages, and N is an integer greater than 1 (106). Thus, when there is only speech data, in a single language, of a target user, multilingual text can be converted into target speech data, in corresponding languages, of the target user, thereby achieving multilingual speech conversion.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zheng, Hong
Abstract
A speech recognition method and a device (800) for speech recognition. The method specifically comprises: receiving speech information inputted by a user (101); obtaining a personalized vocabulary of the user, wherein the personalized vocabulary is established according to historical input content generated according to a process during which the user uses an input method (102); according to the personalized vocabulary, determining a decoding path weight corresponding to the speech information (103); and according to the decoding path weight, determining a speech recognition result corresponding to the speech information (104). The described method and device (800) may improve the accuracy of speech recognition.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Chen, Xiaomin
Zhang, Jingjing
Chen, Wei
Zhao, Chao
Wang, Xiaochuan
Abstract
A speech recognition method and apparatus, and an apparatus for speech recognition. The method specifically comprises: receiving speech data to be subjected to recognition, and determining a first keyword related to the speech data to be subjected to recognition (201); determining a second keyword related to the first keyword according to a knowledge graph (202); and decoding the speech data to be subjected to recognition, adjusting, according to the first keyword and the second keyword, a score of a decoding path corresponding to the speech data to be subjected to recognition, and determining, according to the adjusted score of the decoding path, a speech recognition result corresponding to the speech data to be subjected to recognition (203). According to the method, the accuracy of speech recognition corresponding to keywords relating to an application scenario can thus be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Luo, Dawei
Abstract
An audio data processing method and apparatus, and an electronic device (600, 700) and a storage medium (730). The method comprises: acquiring first multi-channel audio data, wherein the first multi-channel audio data is composed of audio data collected by one or more microphone arrays (101); performing de-reverberation processing on the first multi-channel audio data to obtain second multi-channel audio data (102); generating a time-frequency mask for the second multi-channel audio data (103); performing beamforming processing on the second multi-channel audio data according to the time-frequency mask to obtain first single-channel audio data (104); and outputting an audio signal by using the first single-channel audio data (105). The method realizes audio processing of a plurality of microphone arrays used for non-synchronous collection, thereby preventing the high cost caused by the fact that only unified arrays used for synchronous collection can be used for audio processing, enlarging the pickup range, and improving the robustness.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Yao, Guangchao
Abstract
A data processing method and an apparatus (800) used for data processing. The method specifically comprises: receiving a source language speech input by a target user (101); determining a target acoustic model from a preset acoustic model library according to the source language speech, wherein the preset acoustic model library comprises at least two acoustic models corresponding to different timbre features (102); converting the source language speech to target language speech according to the target acoustic model (103); and outputting the target language speech (104). The method can improve the degree of differentiation of a speaker corresponding to the target language speech output by a translation device, and can enhance the effect of communications of users.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Luo, Dawei
Abstract
Disclosed in embodiments of the present application are a sound collection method and apparatus. Specifically, a microphone array first acquires position information of a user collected in real time from a visual sensor system so as to determine according to the position information of the user a collection direction corresponding to the user. Directional radio sound reception is then performed on the collection direction corresponding to the user. If a target sound signal is received in the collection direction corresponding to the user, then the collection direction in which the target sound signal is received is determined to be a target sound source direction, and then sound collection is performed in the target sound source direction, thereby obtaining a required sound signal. That is, in the embodiments of the present application, a plurality of possible collection directions may be determined and a final target sound source direction may be determined with the aid of the visual sensor system, so that sound collection is performed according to the known sound source direction, which prevents scanning and collection in all directions in space, and improves the accuracy and efficiency of collection.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Ma, Hailei
Abstract
Provided in embodiments of the present invention are an information switching method and apparatus, and an electronic device, wherein the method comprises: acquiring tilt information after the tilt direction of a device is shifted; searching for a preset tilt direction that matches the tilt information, and determining preset information corresponding to the matching preset tilt direction; and switching first input information of the device to second input information, the second input information being determined according to preset information of the matching preset tilt direction. Compared to the prior art requiring manual switching, the embodiments of the present invention may automatically switch the input information according to the tilt direction of the device, which improves the efficiency of information switching.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Sun, Zhiyu
Walayiding, Wailibieke
Zhang, Peng
Abstract
Embodiments of the present application provide a speech acquisition device, comprising a housing, a microphone assembly placed in the housing, and a seal structure located between the microphone assembly and the inner side of the housing. The microphone assembly comprises a microphone body and a microphone circuit board; the seal structure comprises a dust screen and a first adhesive layer; one side of the microphone circuit board is connected to the dust screen; the microphone circuit board and the dust screen are both adhered to the inner side of the housing by means of the first adhesive layer; the other side of the microphone circuit board is connected to the microphone body; the microphone body is provided with a sound hole; the microphone circuit board is provided with a first through hole corresponding to the sound hole; the housing is provided with a second through hole corresponding to the first through hole; the dust screen is disposed between the first through hole and the second through hole; the first through hole, the second through hole, and the sound hole form a sound pickup main channel, and the region outside the sound pickup main channel is sealed. The sealing effect is good.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Liu, Wenwen
Li, Lin
Pan, Muye
Abstract
Provided in the embodiments of the present application are a method and device for data processing and a device for use in data processing. The method specifically comprises: determining information to be replied from clipboard information; the information to be replied corresponding to the entirety or a portion of the clipboard information; determining a reply candidate corresponding to the information to be replied; and displaying the reply candidate. The embodiments of the present application increase reply efficiency and implement smart replying in a scenario of replying across communication windows.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Chen, Lin
Wang, Lu
Liu, Yezhou
Abstract
A data processing method and device, and a device for processing data. Said method comprises: according to a trigger text from a user in a trigger interface, determining a paragraph text corresponding to the trigger text, the trigger interface comprising: a target language translation result display interface corresponding to a source language text (101); and displaying, in the trigger interface, a target text corresponding to the paragraph text, the paragraph text and the target text having a translation relationship (102). Said method can improve the efficiency of a user viewing a source language text and a target language text by contrast.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Jiang, Liyang
Wang, Yuguang
Yang, Jiajun
Shi, Liangliang
Wei, Linyu
Chen, Wei
Abstract
Provided in the embodiments of the present application are a data processing method and apparatus, and an apparatus used for data processing. The method specifically comprises: receiving a speech signal of a first language inputted by a user; determining a speech translation result corresponding to the speech signal and quality information corresponding to the speech translation result, the speech translation result corresponding to a second language; and outputting the speech translation result and the quality information corresponding to the speech translation result. The embodiments of the present application can provide the degree of confidence of a speech translation result to the user, such that the user can objectively understand the accuracy of the speech translation result.
G10L 25/60 - Speech or voice analysis techniques not restricted to a single one of groups specially adapted for particular use for comparison or discrimination for measuring the quality of voice signals
G10L 25/48 - Speech or voice analysis techniques not restricted to a single one of groups specially adapted for particular use
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Han, Qin
Abstract
Provided are an input method and apparatus, and an electronic device, wherein the method comprises: an input method acquiring information of a track drawn in a keyboard region thereof, and displaying the corresponding drawn image according to the information of the track drawn (102); recognizing at least one associated image corresponding to the drawn image (104); and displaying the associated image (106). According to the method, when a user is not satisfied with the drawn image, an image complying with the intention may be selected from the image associated with the drawn image for input without redrawing, and the efficiency of inputting an image is improved.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
G06F 3/0488 - Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser using a touch-screen or digitiser, e.g. input of commands through traced gestures
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Liu, Yezhou
Abstract
Disclosed are a content processing method and device. The processing method comprises: in response to the selection operation of a text, displaying a function option control, wherein the function option control comprises at least one operation option; in response to the triggering of any one operation option, generating a multimedia file that comprises the text; and according to a triggered operation option, transmitting the multimedia file to a target object and/or saving the multimedia file. Therefore, when a user cannot copy, share or save the text, the present invention realizes the sharing or saving of the text by converting the text that needs to be shared or saved into the multimedia file, then sharing the multimedia file to the target object, and/or saving in a local terminal.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Chen, Xiaoshuai
Zang, Jiaojiao
Zhang, Yang
Abstract
An input prediction method and device. The method comprises: acquiring a first input content and a second input content (S101); determining a scenario category corresponding to the first input content (S102); and predicting and acquiring, according to the first input content, the scenario category, and the second input content, a third input content that meets a preset condition concerning a partial order relationship with the first input content, wherein the partial order relationship is used to describe a sequence in which each input content appears (S103). The invention improves accuracy of input prediction, and effectively improves input efficiency.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
45.
DATA PROCESSING METHOD, APPARATUS AND ELECTRONIC DEVICE
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Tu, Chang
Zhou, Weihao
Abstract
Provided are a data processing method, an apparatus and an electronic device. The method comprises: obtaining input information, extracting multiple types of keywords corresponding to the input information (102), wherein the type of the keywords corresponds to the type of multimedia attribute information; matching different types of the keywords, and obtaining corresponding multimedia data according to a result of the matching (104); and recommending the multimedia data (106). Therefore, the input information whose intention cannot be identified can be matched with the corresponding multimedia data and be recommended, and the accuracy of the recommendation information is improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Gao, Shuang
Yu, Hao
Zhang, Tingting
Feng, Ke
Zhao, Bo
Zhang, Long
Liu, Hu
Wang, Xun
Abstract
The present application provides a searching method and related device. The method comprises: turning on a camera device to shoot a real-world scene and generating a virtual object; displaying the real-world scene and the virtual object in a user interface using AR technology; acquiring voice data input by a user and converting the voice data into corresponding text data; sending the text data to a searching server and receiving a searching result corresponding to the text data returned from the searching server; and in the user interface which displays the real-world scene and the virtual object, displaying text data corresponding to the searching result using the AR technology, and/or outputting voice data converted from the searching result. Thus, according to the embodiment of the present invention, a plurality of searching results can be more intuitively obtained for viewing, thereby improving searching efficiency; besides, the sense of the real-world scene is enhanced by means of voice interaction with the virtual object and the user experience is thus improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Liu, Shiwen
Abstract
Disclosed in embodiments of the present application are an expression input method and apparatus. The method specifically comprises: displaying M candidate words on the basis of the input operation of a user, M being a positive integer; and displaying, on the basis of the trigger operation of the user, expressions corresponding to all or some of the M candidate words. The present application can solve the technical problem in the prior art of inconvenience in expression input, can improve the convenience of expression input, and can increase the input speed of a user.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xiao, Ziqiu
Zhang, Jie
Sun, Peining
Abstract
Provided is a translation pen, including a rear sleeve, a front sleeve, a top cover assembly, a bottom cover assembly and a main body assembly. The rear sleeve and the front sleeve are hollow cylindrical structure, and the hollow portions of the rear sleeve and the front sleeve are connected as a receiving chamber after the rear sleeve is embedded into the front sleeve. The main body assembly is received in the receiving chamber. The end of the main body assembly adjacent to the top cover assembly is threadedly connected to the top cover assembly, the end of the main body assembly adjacent to the bottom cover assembly is threadedly connected to the bottom cover assembly, and the front and rear sleeves can be tightly locked by threaded connection of the top cover assembly and the bottom cover assembly to the main body assembly. The present invention solves the problem of the insufficient structural strength caused by the connection between the front sleeve and the rear sleeve through cartridge modes such as buckles, guarantees the structural strength of the translation pen, and effectively protects the main body assembly inside the front sleeve and the rear sleeve.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xu, Jingfang
Zhang, Jiacheng
Zhai, Feifei
Dai, Lei
Liu, Qi
Fang, Fei
Ding, Yanzhuo
Abstract
Provided are a cross-language search method and apparatus, and an apparatus for a cross-language search. The method specifically comprises: translating a source language search term into a target language search term; with respect to the target language search term, acquiring a target language search result and a source language translation result, corresponding thereto, wherein the source language translation result is obtained by translating the target language search result; acquiring a keyword matching the target language search term from the source language translation result; and marking and displaying the keyword in the source language translation result. The embodiments of the present application can improve a coverage ratio of content marked in red in a source language translation result corresponding to a target language search term, and thus can improve the efficiency of acquiring information from the source language translation result corresponding to the target language search term.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
TSINGHUA UNIVERSITY (China)
Inventor
Xu, Jingfang
Liu, Yang
Zhang, Jiacheng
Zhai, Feifei
Dai, Lei
Liu, Qi
Fang, Fei
Ding, Yanzhuo
Abstract
Provided are a cross-language search method and apparatus, and an apparatus for a cross-language search. The method specifically comprises: translating a source language search term into a target language search term; with respect to the target language search term, acquiring a target language search result and a translation search result, corresponding thereto, wherein the translation search result is obtained by translating the target language search result; acquiring a keyword matching the target language search term from the translation search result; and replacing the keyword in the translation search result with the source language search term. The embodiments of the present application can improve the accuracy of a translation search result and the consistency between the translation search result and a target language search result, and can also improve the attraction of the translation search result for a user, so that the click rate of the translation search result can be further improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xiao, Ziqiu
Zhang, Jie
Sun, Peining
Abstract
Provided is a translation machine, comprising a housing, a microphone assembly, a speaker, and a switch button; the housing comprises a bottom and a frame; the bottom and the frame constitute an accommodating chamber; one frame of two adjacent frames is provided with the switch button, and the other frame is provided with a first sound-pickup hole of the microphone assembly and a first sound outlet hole of the speaker; the microphone component and the speaker are located in the accommodating chamber, and are arranged facing the first sound-pickup hole and the first sound outlet hole; the microphone assembly consists of two microphones, and the speaker is located between the two microphones. The two microphones and the speaker are located at the same end of the translation machine and during translation, the microphone and speakers operate alternatingly; therefore, the speaker and the microphones do not affect one another; at the same time, two microphones are arranged, such that the directivity of sound pickup is enhanced and sensitivity is high, improving the accuracy of speech translation results; furthermore, it easy to align the speaker and microphone to both parties to a conversation and operability is very good.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xiao, Ziqiu
Zhang, Jie
Sun, Peining
Abstract
A translation machine, comprising a housing (100), a display screen (200), and switch buttons (300). The display screen (200) is arranged on the housing (100). The housing (100) and a first edge frame (10) adjacent to the top of the translation machine are provided with a protruding part (11) protruding towards the exterior of the first edge frame (10). The protruding part (11) is provided in proximity to the top of the translation machine. The switch buttons (300) are arranged on the protruding part (11). Because the switch buttons (300) are arranged at positions on the edge frame of the translation machine in proximity to the top of the translation machine, when the translation machine is held by hand, the protruding part (11) is located precisely at the position of the thumb of the person holding, and, with the protruding part (11) in proximity to the position of the thumb, switching between translation languages can be performed quickly and conveniently with the thumb, thus making the operation of the translation machine to further comply with habits of the human body, increasing translation efficiency, and providing great operability.
G06F 15/02 - Digital computers in general; Data processing equipment in general manually operated with input through keyboard and computation using a built-in program, e.g. pocket calculators
G06F 17/28 - Processing or translating of natural language
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xiao, Ziqiu
Zhang, Jie
Sun, Peining
Abstract
A mainboard and a translation machine. The mainboard comprises a body, and a shielding area and a non-shielding area on the front surface of the body; the shielding area and the non-shielding area are provided with devices connected to the body, wherein the height of a device of the maximum height in the shielding area is smaller than the height of a device of the maximum height in the non-shielding area; since the device of a relatively large height is disposed in the non-shielding area, the height of the device in the shielding area is relatively small, and after a shielding cover is fixed in the shielding area, the height of the shielding cover can be effectively reduced, thereby reducing the height of the entire mainboard, reducing the thickness of the translation machine, and facilitating the lightening and thinning development of the translation machine.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Tu, Chang
Zhang, Yang
Wang, Yanfeng
Abstract
Disclosed in embodiments of the present application are a candidate word prediction method and device. In the method, according to content most recently input on the screen, on-screen content requiring word prediction can be determined; since a word prediction intention can reflect, to a large extent, a user's actual input intention for the on-screen content, if the word prediction intention of the on-screen content can be determined, content that may be continuously input on the screen by the user after the on-screen content can be determined according to the word prediction intention. Therefore, when data corresponding to the word prediction intention is obtained according to a keyword in the on-screen content, and a candidate corresponding to the on-screen content is predicted according to the data, the candidate determined according to the data is more likely to meet the user's input intention. Thus, the input convenience for the user is improved. Moreover, the determined candidate does not need to be generated on the basis of the past input data of the user, and thus, the likelihood and accuracy of the predicted candidate are improved.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
G06F 17/30 - Information retrieval; Database structures therefor
55.
VIDEO PROCESSING METHOD AND DEVICE, AND DEVICE FOR VIDEO PROCESSING
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhang, Jie
Bu, Hailiang
Jin, Yixiao
Xing, Zhenzhen
Jiang, Pin
Feng, Xinqiang
Abstract
The embodiments of the present application provide a video processing method and device, and a device for video processing, the method particularly comprising: recognizing a video stream and/or an audio stream corresponding to a video, to obtain a corresponding recognition result; acquiring, from a preset item database, a target item matching the recognition result; and adding target information corresponding to the target item into a video frame corresponding to the video stream and/or the audio stream. The embodiments of the present application can shorten video processing time and improve video processing efficiency, and can improve video coverage of target information.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xu, Meng
Han, Qin
Li, Lin
Cai, Pengli
Chen, Zhangqin
Abstract
Embodiments of the present application provide an input method, device, and machine-readable medium. The input method specifically comprises: obtaining the data of movement inputted by a user on an input interface; identifying the movement data so as to obtain the target character and/or target image corresponding to the movement data; displaying the target character and/or target image as a candidate. The embodiments of the present application effectively reduces user operation path and enhances user input efficiency.
G06F 3/0488 - Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser using a touch-screen or digitiser, e.g. input of commands through traced gestures
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xiao, Ziqiu
Zhang, Jie
Sun, Peining
Abstract
A watch clasp and a wearable apparatus, the watch clasp specifically having: a clasp frame (1), wherein, the clasp frame (1) has a semicircular or circular, and a reinforcement shaft (2) is provided between two opposing edges of the clasp frame (1). The present application can, by means of the reinforcement shaft (2), increase the strength of the watch clasp, and so can to a certain extent avoid deformation of the watch clasp when in use.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Mingxiu
Yin, Lei
Bu, Hailiang
Abstract
A voice-based data processing method and apparatus, and an electronic device, used for completely recording a diagnostic inquiry process. The method comprises: acquiring diagnostic inquiry process data, the diagnostic inquiry process data being determined according to voice data collected in a diagnostic inquiry process (102); performing recognition according to the diagnostic inquiry process data to acquire corresponding first text data and second text data, wherein the first text data belongs to a target user, and the second text data belongs to another user other than the target user (104); and obtaining diagnostic inquiry information according to the first text data and the second text data (106). By means of the method, the apparatus and the electronic device, the statements of a doctor and a patient in a diagnostic inquiry process can be automatically distinguished, the diagnostic inquiry process is completely recorded, and same is automatically organised to obtain content such as a medical record, thus saving on time for organising diagnostic inquiry records.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xiao, Ziqiu
Zhang, Jie
Sun, Peining
Abstract
A watch body comprising: a primary watch body (2) and a secondary watch body (2), where the left lateral surface or the right lateral surface of the secondary watch body (1) is fused and connected with the primary watch body (2), and the secondary watch body (1) is provided with a first camera (3) and a second camera (4). The watch body allows a user to implement the demand for photographing anytime and anywhere via a wearable device. This allows the stacking of the watch body corresponding to a movement to be of an increased degree of rationality and the spatial arrangement to be of increased compactness; therefore, the size of the watch body corresponding to the movement can be reduced as much as possible. Also provided is the wearable device comprising the watch body.
G04G 21/00 - Input or output devices integrated in time-pieces
G03B 29/00 - Combinations of cameras, projectors or photographic printing apparatus with non-photographic non-optical apparatus, e.g. clocks or weapons; Cameras having the shape of other objects
G04B 47/00 - Time-pieces combined with other articles which do not interfere with the running or the time-keeping of the time-piece
60.
SEARCH METHOD AND APPARATUS, AND APPARATUS FOR SEARCHING
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Junhao
Lin, Jiansu
Abstract
Provided are a search method and apparatus, and an apparatus for searching, wherein the method specifically comprises: acquiring a search word with a meaningless target corresponding to the current search word input by a user; acquiring a relevant search result corresponding to the search word with a meaningless target; and displaying the relevant search result as a first search result item corresponding to the search word. In the embodiments of the present application, pre-set information related to a search word with a meaningless target can be carried by means of a relevant search result, so that the search experience of a user can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Mansuer, Maerhufu
Cui, Xin
Zhang, Yang
Abstract
Provided are a text editing method and device, and an electronic apparatus. The method comprises: acquiring a text to be edited (S101); upon receiving a trigger operation, generating an edit region (S102); segmenting the text into N text segments, and separately displaying the N text segments in the edit region, where N is a positive integer greater than 1 (S103); and upon receiving an edit operation with respect to the N text segments, editing the N text segments according to the edit operation (S104). The method and device solve a technical problem in which a text editing process requires complex operations and has low editing efficiency, thereby improving text editing efficiency.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xiao, Ziqiu
Sun, Peining
Zhang, Jie
Abstract
Disclosed is a smart watch. The smart watch comprises a casing (1), wherein a SIM card port (3) and a charging port (2) are provided on the casing (1), the SIM card port (3) and the charging port (2) are provided on the same side wall of the casing (1), and the smart watch further comprises a protective cover (4), the protective cover (4) covering the outer side of the SIM card port (3) and the charging port (2).When worn normally, the protective cover (4) isolates the SIM card port (3) and the charging port (2) from the external environment, preventing contaminants in the external environment from entering the SIM card port (3) and the charging port (2), and thus protecting the electronic elements inside the casing (1). When a SIM card is plugged in or out or when charging, the protective cover (4) can be opened to expose the SIM card port (3) or the charging port (2). The protective cover (4) is designed as relatively independent first protective cover (41) and second protective cover (42), and the first protective cover (41) and second protective cover (42) can be independently opened or closed, so that when charging is required, only the second protective cover (42) needs to be opened.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Fupo
Liu, Zhankun
Wang, Ying
Abstract
The embodiment of the present application relates to the field of communications and provided are a contact method, an electronic device, and a server, so as to solve the technical problem of high cost of the electronic device caused by a function of adding a contact in the prior art. The method comprises receiving a first contact adding request associated with a first sound signal sent by a second electronic device, the first sound signal corresponding to first identification information of a first electronic device, and the second electronic device obtaining the first sound signal played by the first electronic device by means of an answering manner; determining the first identification information and the second identification information of the second electronic device on the basis of the first contact adding request; and adding first user of the first electronic device as the contact of the second user of the second electronic device on the basis of the first identification information and the second identification information. In the embodiment of the present application, the efficiency of adding contacts is improved.
H04M 1/2745 - Devices whereby a plurality of signals may be stored simultaneously with provision for storing more than one subscriber number at a time using static electronic memories, e.g. chips
64.
METHOD, DEVICE AND EQUIPMENT FOR PROCESSING MEDICAL RECORD INFORMATION
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Yin, Lei
Li, Mingxiu
Pu, Hailiang
Wei, Shijia
Abstract
Disclosed by the present invention is a method for processing medical record information. The method comprises: acquiring an original medical record text and dividing the original medical record text into at least one target text unit; determining a target category corresponding to a text feature of the target text unit; and generating a target medical record text, the target text unit in the target medical record text being embodied as text information under the target category. By means of the method provided by the embodiments of the present invention, different information content in a structured target medical record text is respectively divided into corresponding categories, which not only may enable a user to read more smoothly, but may also enable the user to find desired information content more quickly so that the target medical record text is more favorable for data organization and analysis. In addition, also disclosed by the present invention are a device and equipment for processing medical record information.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Chen, Xiaoshuai
Maerhufu, Mansuer
Zhang, Yang
Abstract
An input method, a device, and an electronic device (800), the method comprises: obtaining input operations (S101); determining input contents corresponding to the input operations (S102); determining, on the basis of at least a keyword included in an adaptive words library, an input candidate matching the input content, wherein the keywords included in the adaptive words library are determined on the basis of current page contents of the electronic device (800) (S103). With this method, the technical effect of providing more accurate input candidates can be achieved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhang, Yang
Jiao, Haofeng
E, Yanli
Abstract
An input method and apparatus, and an apparatus for an input. The method specifically comprises: receiving an input string having a fast input intention, wherein the fast input intention is used for inputting, by means of the input string corresponding to brief information about a word or a word group, the word or the word group (101); according to a statistical language model, acquiring a word candidate and/or a word group candidate corresponding to the input string, wherein the word candidate and the word group candidate are respectively a complete word and a complete word group corresponding to the input string (102); and displaying the word candidate and/or the word group candidate to a user (103). The method can improve the flexibility and application scope of fast input, and can improve the quality of a word candidate and/or a word group candidate, thereby improving input efficiency.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
67.
SEARCH METHOD AND APPARATUS, AND ELECTRONIC DEVICE
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xu, Jingfang
Han, Yifan
Ru, Liyun
Wang, Xiaochuan
Abstract
The present application relates to the field of the Internet. Disclosed are a search method and apparatus and an electronic device, for use in resolving the technical problem in the prior art of less search results provided when a search is performed in a cross-language manner. The method comprises: obtaining a search key word; providing at least two pieces of translation obtained when the search key word is translated; and obtaining a translation result obtained when a search is performed by means of the at least two pieces of translation. The technical effect of increasing search results obtained when a cross-language search is performed is achieved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Yin, Lei
Li, Mingxiu
Bu, Hailiang
Abstract
Provided are a prescription matching method and device, and a device for prescription matching. The prescription matching method specifically comprises: acquiring a drug vocabulary set to be matched; determining a spatial vector corresponding to each item of drug vocabulary in the drug vocabulary set, constituting a spatial vector group corresponding to the drug vocabulary set; using a prescription matching model to perform drug matching on a specified attributive character of drug vocabulary corresponding to the drug vocabulary set, so as to obtain a matching prescription matched with the specified attributive character; wherein the specified attributive character comprises: the spatial vector group; the prescription matching model being obtained by performing training on drug attributive characters corresponding to a prescription sample. According to the embodiments of the present application, performing the drug matching according to more plentiful information can improve the accuracy of the drug matching.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Lin, Jiansu
Abstract
An information processing method and apparatus based on a search engine. The method applied to a first terminal specifically comprises: receiving a search word input by a first user of a first terminal (101); when the search word satisfies a pre-set question condition, acquiring a question request corresponding to the search word (102); and sending the question request to a second terminal corresponding to a second user, wherein position information about the second terminal matches position information corresponding to the question request (103). By means of the method, the timeliness and accuracy of answering a question can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Hou, Baicen
Zhu, Fangfang
Abstract
Disclosed are a method and a device for matching an image to search information. The method comprises: in response to a search request of a user, acquiring a first search result corresponding to a search keyword comprised in the search request (S201); if it is determined that the first search result does not comprise an image, searching, according to the search keyword, a pre-established first image library, and acquiring an image matching the search keyword, wherein the first image library stores a correspondence between an image and a search keyword (S202); associating the image matching the search keyword with the first search result and displaying the image and the first search result (S203). The method of the present invention automatically matches a suitable image to search information, thereby improving the readability of information, allowing users to acquire information more intuitively, reducing frequency of user/network interaction and reducing wastage of network resources.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Yubin
Zhao, Xuan
Han, Yifan
Xie, Yuanjiang
Abstract
A searching method, searching device and searching equipment. The method comprises: obtaining first feature information which is information in a first language (201); converting the first feature information into second feature information in response to a search request in which the first feature information belongs to a first search state (202), wherein the search request in the first search state is used for instructing the feature information converted on the basis of a designated language to search a search result, and the second feature information is information of a second language corresponding to the first feature information; and searching a search result matched with the second feature information and feeding the search result back (203). By the method provided, the operation for performing search in an unfamiliar language can be simplified and therefore the user experience of search service can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Lin, Jiansu
Abstract
Provided are a search method and device, and device used for searching; the search method comprises a preprocessing process and a search process, the preprocessing process comprising: obtaining target data (101), and establishing for each search context a mapping relationship from master data to sub-data in each context (102); said master data and said sub-data belonging to said target data and the sub-data being associated with the master data; said search process comprising: responding to a search request triggered by a user, and obtaining an initial keyword related to said search request (103); according to the master data corresponding to said initial keyword and the mapping relationship with said master data, searching to obtain slave data associated with said master data (104); combining the master data and slave data to obtain target search data (105). Using the search method and device improves search efficiency and improves user search experience.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Huabei
Zhu, Fangfang
Tian, Xiaona
Tang, Zhong
Hou, Baicen
Abstract
An information pushing method and apparatus, and an electronic device, relating to the technical field of Internet, and for use in solving the technical problem in the prior art that personalized recommendation cannot be provided during network search. The method comprises: receiving a search keyword by means of a search box of a search engine (S101); acquiring a recommended associated word corresponding to the search keyword (S102); and providing a current user with the recommended associated word and recommendation information that is preset by a particular user, wherein the recommendation information is associated with the search keyword and/or the recommended associated word (S103). The technical effect of personalized recommendation on the basis of customization of a user is achieved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Lin, Changqing
Abstract
A key processing method and apparatus. The method specifically comprises: in response to a trigger operation of a user on a current key in a current interface, processing a non-trigger state key, wherein the non-trigger state key is a key apart from the current key in the current interface (101); and outputting a processing result of the non-trigger state key (102). The method and apparatus not only can enhance the richness of a current interface, but also can enhance the feedback effect and user experience in the process of using the interface, and can improve the accuracy of a key operation.
G06F 3/0484 - Interaction techniques based on graphical user interfaces [GUI] for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Liping
Wang, Suhang
Yan, Congxian
Yang, Lei
Liu, Min
Zhao, Hong
Yao, Jia
Abstract
Provided are a speech input method and a terminal device. The method comprises: in a speech input mode, receiving a first speech item inputted by a user, recognizing the same to generate a first recognition result, and displaying, according to the first recognition result, corresponding text to the user (S11); in an editing mode, receiving a second speech item inputted by the user, and recognizing the same to generate a second recognition result (S12); and converting the second recognition result into an editing instruction, and performing a corresponding operation according to the editing instruction, wherein a switch can be performed between the speech input mode and the editing mode (S13). The invention enables a speech input to be performed in the speech input mode or the editing mode, and a switch between the speech input mode and the editing mode not only realizes conversation of the speech input into text but also enables a corresponding editing operation to be performed according to the speech input from a user, thereby increasing speech input efficiency and enjoyment, and improving user experience.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Yao, Guangchao
Abstract
The present application relates to the field of audio technology, provided in embodiments are a method and device for audio identification. The method comprises: capturing an audio stream of a first length from source data of a video file; retrieving corresponding audio information according to the audio stream of a first length, and displaying the same to a user; the corresponding audio information retrieved according to the audio stream of a first length comprising: dividing the audio stream into at least two sub-audio streams according to a preset rule; sequentially retrieving the sub-audio streams obtained from the dividing, so as to obtain the audio information. The embodiments of the present application are able to extract an audio stream directly from a video source data which is currently being played so as to be retrieved, which requires no extra recording operations and is unaffected by noisy environments. The present application has simple operations and is highly accurate; the retrieving process thereof would not affect a user's normal viewing a video, while improving retrieving efficiency and success rate.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Sun, Jian
Li, Yi
Abstract
Disclosed are a user proxy self-adaptation uniform resource locator prefix mining method and device. The method comprises: collecting a website homepage grabbed by a web crawler of a search engine on a computer side; judging whether the website homepage is a UA self-adaptation webpage; judging whether each first-order prefix corresponding to a web address included in a website meets the UA self-adaptation condition when the website homepage is a UA self-adaptation webpage; and determining that the first-order prefix meeting the UA self-adaptation condition is a UA self-adaptation URL prefix. By means of the embodiments of the present application, URL prefixes meeting the UA self-adaptation condition included in various common websites can be mined out. On this basis, UA self-adaptation webpages can be effectively removed during the operation of transformational rule mining, and the defect of increasing workload or obtaining an erroneous transformational rule during transformational rule mining of UA self-adaptation URL in the prior art is avoided.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Xing, Zhenzhen
Bu, Hailiang
He, Li
Abstract
An information processing method and apparatus, and an apparatus for information processing. The information processing method specifically comprises: acquiring a food picture of a user and geographical location information corresponding to the food picture (101); determining a dining room corresponding to the geographical location information (102); searching in dish pictures corresponding to the dining room to obtain a result dish picture which is the same as or similar to the food picture (103); according to a dish name corresponding to the result dish picture, determining a dish name corresponding to the food picture (104); and outputting the dish name corresponding to the food picture (105). According to the method and apparatus, difficulties in information acquisition can be reduced, and the efficiency of information acquisition can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Jian
Mao, Hong
Abstract
A fuzzy keyword matching method and apparatus. The method comprises: acquiring a keyword to which each character in a to-be-matched text pertains, and an index bit of the character in the pertaining keyword according to a keyword set respectively (101); judging the index bit of the character in the pertaining keyword (102); if the character is a first character, recording the pertaining keyword of the character, and marking the first character in the keyword into the to-be-matched text (103); and if the character is not the first character and when recorded keywords exist, searching the recorded keywords for the pertaining keyword of the character, and marking the character in the keyword into the to-be-matched text (104). When each character in a keyword is marked into the to-be-matched text, it is determined that the to-be-matched text hits the keyword. The method and apparatus can overcome the defect in the prior art of low keyword matching efficiency to a certain extent.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Tu, Chang
Zhang, Yang
Wang, Yanfeng
Abstract
Provided in an embodiment of the application are a recommendation method and device, and a device for recommendation. The recommendation method specifically comprises: after an input method program has been triggered, acquiring current environment information in which the input method program is involved; analyzing historical input behavior data in historical environment information associated with the current user, so as to acquire a target entry matching the current environment information from the historical input behavior data; and displaying the target entry. The embodiment of the present invention can provide a candidate option to a user before the user performs any input operation, thereby significantly improving input efficiency of the user.
G06F 17/30 - Information retrieval; Database structures therefor
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Lin, Changqing
Abstract
An input method and apparatus. The method specifically comprises: displaying an input interface, wherein the input interface comprises at least one button (101); and responding to a trigger operation performed by a user on the button during an input process, and playing audio that matches the button and satisfies the characteristics of the input interface (102). The method can feed back an animation effect and dynamic audio that satisfy a current input interface scene to a user according to a button pressed by the user during an input process, and can also enrich button pressing tones during the input process.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
82.
INPUT METHOD, INPUT DEVICE, SERVER AND INPUT SYSTEM
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Liu, Hui
Abstract
An input method, input device, server and input system. The method specifically comprises: generating a three-dimensional space (101); detecting an operating position sequence generated by a user in the three-dimensional space (102); determining button regions corresponding to the operating position sequence (103); determining, according to the button regions, a button sequence corresponding to the operating position sequence (104); and performing, according to the button sequence, conversion to obtain a corresponding candidate item (105). The above method can improve flexibility and convenience of text input.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Hao, Qianqian
Mansuer, Maerhufu
Cui, Xin
Zhang, Yang
Abstract
Provided is an information processing method and device, which simplify the operation flow. The method comprises: acquiring interaction information sent by an opposite end in the process of conducting interaction in a preset communication manner; determining a preset trigger condition matching the interaction information through analysis, so as to acquire corresponding display information; and recommending the display information corresponding to the matched preset trigger condition. The information which a user may be interested in is directly recommended and displayed in the process of interaction, thereby simplifying the operation flow and improving the efficiency.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Qiao, Yujing
Chen, Bin
Wang, Dong
Yu, Hao
Zhang, Kuo
Abstract
An information input method and apparatus, which relate to the technical field of input methods. The method comprises: obtaining candidate items corresponding to an input code string, and determining a target search string from the candidate items (110); searching for the target search string, and obtaining a corresponding search result (120); extracting corresponding target content from the search result (130); and when it is detected that an information sending interface of a first chat client side is triggered, reconstructing the target content into first template information capable of being recognized by the first chat client side by adopting a first template provided by the first chat client side, and sending the first template information to a second chat client side (140). The method ensures a coherent chat process of users, has no any redundant switching operations and basically has no switching price and cost, the users are not required to perform complicated copy/paste, screenshot or share processes to send the search result to a chat peer, and accurate and concise target content can be automatically organized from the search result so as to be sent to a target client side.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Li, Yanmei
Li, Ping
Abstract
An information presentation method based on input and an input method system solve the problem of low efficiency due to separately inputting words and expressions in an input method. The method comprises: adopting context information to identify a received input character string, to identify a current semantic keyword (102); determining presentation information matching the semantic keyword, and presenting the presentation information in the form of a pop-up bubble (104), wherein the presentation information comprises but is not limited to an expression picture; and displaying the presentation information according to a user instruction (106), wherein the displaying comprises but is not limited to performing an on-screen operation on the expression picture. A current semantic keyword identified by the method and device is not limited to a candidate item corresponding to an input character string, and semantic identification based on context information better meets an input intention of a user and presentation information also better meets the input intention of the user, thereby improving the input efficiency of presentation information, such as words and expressions.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Xiaofei
Liu, Min
Du, Conghui
Yang, Lei
Li, Liping
Abstract
A method for adjusting an input-method virtual keyboard and an input-method device, which are used to solve the problem that an existing input-method virtual keyboard is difficult to operate with one hand. The method comprises: adjusting the size of an input-method window to a screen size according to a received keyboard floating instruction; intercepting a click event performed on a touchscreen by a user and detecting the click event; and when detecting that the click event is a keyboard adjustment instruction, adjusting the drawing position and/or drawing size of an input-method view according to the keyboard adjustment instruction, the input-method view being a region for drawing a virtual keyboard in the input-method window. Since the entire screen region is the input-method window, the input-method view for drawing the virtual keyboard floats in the input-method window and the position and size of the input-method view can be adjusted randomly, so as to facilitate one-hand operation of the user, realize rapid input and improve the input efficiency.
G06F 3/0484 - Interaction techniques based on graphical user interfaces [GUI] for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wang, Yi
Zhao, Hong
Abstract
A method and device for inputting English. The method comprises: receiving an input English character string (101), the English character string having the original length; calculating the target length according to the original length of the English character string (102); searching for a target English word of which the length is between the original length and the target length (103); calculating the maximum common length of the target English word and the English character string (104), the common length being the length of the same letters; and returning the target English word when the maximum common length is N times the original length (105), N being a positive number less than or equal to 1. In this way, when a user forgets the spelling of an English word or in case of few inputs, the required English word can be obtained by inputting an English letter string that is part of the English word, so the convenience in operation is greatly improved.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Cui, Xin
Ren, Shangkun
Tang, Zheng
Zhang, Yang
Abstract
An input method and apparatus and an electronic device. The input method comprises: acquiring text information at an input cursor (101), wherein the text information comprises upper-part text information before the input cursor and/or lower-part text information after the input cursor; extracting a key word in the text information (102); searching an associated candidate word bank of the key word, and acquiring an on-screen candidate word queue at the input cursor (103); and outputting the on-screen candidate word queue (104). By acquiring the text information at the input cursor, and determining the on-screen candidate word queue on the basis of the key word in the text information, the problem in the prior art that after the position of the input cursor is changed, a reliable on-screen entry cannot be acquired and the on-screen candidate word cannot be associated and provided can be solved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Cui, Xin
Ren, Shangkun
Su, Qilong
Zhang, Yang
Abstract
An input method and apparatus and an electronic device. The input method comprises: acquiring historical dialogue information about a dialogue window (101); extracting characteristic information in the historical dialogue information, wherein the characteristic information comprises a key word (102); determining prediction information, wherein the prediction information comprises the key word (103); acquiring an on-screen candidate word queue in an input text box in the dialogue window according to the prediction information (104); and outputting the on-screen candidate word queue (105). By fully excavating the historical dialogue information about the dialogue window, determining the prediction information by integrating the characteristic information extracted from the historical dialogue information, and searching a corresponding associated candidate word bank according to the prediction information, the final on-screen candidate queue is obtained. According to the method, no matter whether the input text box is blank or not, the on-screen candidate queue still can be acquired; moreover, the acquired on-screen candidate queue is obviously more reasonable and more accurate than that acquired only according to an on-screen entry of a user last time.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Ren, Shangkun
Zhang, Yang
Abstract
A text information input method and device, the method comprising: when detecting that all or part of the current on-screen text information is deleted, determining first and second text information (101), the first text information comprising a string of key inputs corresponding to all the on-screen text information, all the on-screen text information, and the number of backspacing times for deletions, and the second text information comprising text information redisplayed on the screen following the appearance of the first text information; according to the first text information, calculating a selection probability of selecting the second text information for display on the screen (102); displaying the second text information in a specified position when the selection probability is greater than a preset probability threshold (103). The method and device can effectively reduce user keying, increase input efficiency and realize quick and simple input, thus improving user experience and reducing consumption of device system resources.
G06F 3/023 - Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
91.
SEMANTIC COMPREHENSION-BASED EXPRESSION INPUT METHOD AND APPARATUS
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Gu, Siyu
Liu, Huasheng
Zhang, Kuo
Abstract
Disclosed are a semantic comprehension-based expression input method and apparatus, which relate to the technical field of input methods. The method comprises: obtaining text content corresponding to an input sequence (110); performing word segmentation on the text content, and extracting various text features on the basis of the result of word segmentation (120); using the text features to construct input vectors, and substituting the input vectors into an emotion classification model to classify the input vectors, in order to determine emotional labels to which the text content belongs (130); obtaining expressions of various themes under the emotional labels respectively on the basis of the corresponding relation between the emotional labels and the expressions of the various themes (140); and sorting the expressions of the various themes, and displaying the expressions on a client as candidate items (150). Thus, input of expressions by a user is facilitated, the expression input efficiency is improved, and abundant and wide-range expression resources are provided for the user.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Sun, Mingze
Ren, Bin
Gao, Zhi
Abstract
A method is disclosed for pushing information to a user of a computer system. The method includes determining network information interested by the user based on at least one of the user's browsing history of first one or more websites and the user's input related to the user's interest. The method also includes determining network resources to be monitored, and dynamically monitoring the network resources determined to be monitored. Further, the method includes determining updated information from monitored network resources related to the user's interest, and pushing the updated information to the user.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Jiang, Youxin
Abstract
A method and system for providing network resources are disclosed. Wherein, the method comprises: identifying an invalid network resource (S201); retrieving an alternate network resource for the invalid network resource by using characteristic information of the invalid network resource (S202); providing the alternate network resource to a subscriber (S203). By this solution, the success rate of providing network resources can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wan, Yong
Abstract
A method and a system for displaying the page element in the web page are disclosed. The method includes: determining the target page element in the web page (S201); and displaying the target page element on the fixed position relative to the window of the browser in which the web page is located (S202). With the method, it is instrumental in improving the coefficient of utilization of the network resource.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhao, Feng
Qi, Tejia
Abstract
A method and device for outputting web form data and a form processing system are disclosed. The method comprises: if a trigger event outputted by a web form universal data set is monitored, then form item data, which is matched with the form item in a web form to be filled, is searched in the web form universal data set which is obtained by analyzing and summarizing form items of at least two stored forms in advance; if the matched form item data is found out in the web form universal data set, then the matched form item data is outputted. The device includes a form processing module, a matching and searching module, and a data output module. The system includes a form searching module and a summarizing and analyzing module. Embodiments of the technical solution can increase the filling efficiency of a web form, making for improving the user experience.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wan, Yong
Abstract
A method and system for showing network resources are disclosed, wherein said method includes the following steps: at least one of network resources needing to be showed independently is determined, and network attribute information of said network resources is obtained; a resource list window is obtained, and the identification information of at least one of said network resources is displayed in said resource list window; when the network resources related to the identification information in said resource list window need to be showed independently, said network attribute information of said network resources is used to show the network resources in an independent showing mode. Through this invention, the efficiency for a user to obtain network resources can be improved.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhang, Yan
Wang, Jian
Cao, Zhen
Abstract
A method and an apparatus for character input using direction operation are disclosed in the present invention. Said method includes: predefining multiple direction operations for an input device; receiving a direction operation sequence inputted by a user, wherein the direction operation sequence includes more than two consecutive direction operations; outputting a character corresponding to said direction operation sequence. The direction operations can be enabled with a stick. The stick is composed of a central stick, several margin sensors, and a down sensor located below the central stick. The central stick is also assembled with a pressure-sensitive switch, which enables the connected or disconnected status of the central stick with each margin sensor or the down sensor. The input method of the input device provided by this invention presents the advantages of user convenience and higher input efficiency.
G06F 3/00 - Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Zhou, Dakui
Gao, Zhi
Abstract
An update notification method and browser. The method includes: the browser monitors network resource dynamically (101); if the update of the network resource is found, the method will notify the user in a preset update notification manner (102). As the update notification method monitors the web page directly via the browser, it is able to monitor various kinds of network resources dynamically including the web page that do not support RSS output, and when the update of the network resource is found, the method will notify the user initiatively and timely in multiple manners.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Gao, Zhi
Wang, Xiaochuan
Abstract
An update notification method and system. The method includes: fetching network resource (301); extracting specific contents of the network resource (302); comparing the specific contents of the network resource extracted twice successively (303); notifying users that the content of the network resource is updated if the specific content changes (304). The method is able to monitor various network resources dynamically including the network resources that do not support RSS output, and notify users after correlative network resource is updated.
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO., LTD. (China)
Inventor
Wu, Jian
Zhao, Feng
Abstract
A method and a browser for loading a webpage in a browser are provided. The browser is a multi-kernel browser. The method includes the following steps: receiving the first request information that a user loads webpage (S201); acquiring adaptive relationship between a webpage requested by the user and the browser kernel according to a preset rule; selecting the corresponding browser kernel to load the webpage requested by the user according to the adaptive relationship. The other three methods and browsers for the loading webpage in the browser are also disclosed. Embodiments of the invention enable the intelligent selection of the kernels in a browser, fully utilize the advantages of each kernel, avoid influence to the loading webpage of the user caused by defects of each kernel, process intelligently advantage complementation between multiple kernels in browsers and enable improving effect of browsing the webpage via the browser to a certain extent.