Abrégé

A system, comprising: obtaining first point cloud data generated by a first scanner device of a first autonomous driving vehicle (ADV). The system applies one or more abnormality detection algorithms to the first point cloud data to determine an abnormality in the first point cloud data. The system determines an abnormality type for the abnormality. The system determines the abnormality is above a severity threshold based on the abnormality type. In response to determining that the abnormality is above the severity threshold, the system indicates a warning to inspect the first scanner device for any hardware malfunctions.

Classes IPC  ?

• B60W 40/10 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés au mouvement du véhicule
• G01S 17/88 - Systèmes lidar, spécialement adaptés pour des applications spécifiques

Abrégé

An autonomous driving vehicle (ADV) processes a set of sensor data concurrently by both a first computing unit and a second computing unit, wherein the set of sensor data is generated by a sensor (810). The first computing unit formats the set of sensor data into a set of message data (812). The second computing unit determines whether the set of sensor data indicates a control path fault (814). The second computing unit reports the control path fault responsive to determining that the set of sensor data indicates the control path fault (816).

Classes IPC  ?

• G05B 19/418 - Commande totale d'usine, c.à d. commande centralisée de plusieurs machines, p.ex. commande numérique directe ou distribuée (DNC), systèmes d'ateliers flexibles (FMS), systèmes de fabrication intégrés (IMS), productique (CIM)
• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions

Abrégé

In one aspect, a computing device of an autonomous driving vehicle (ADV) is configured to determine a first control signal for a light detection and ranging (Lidar) sensor of the ADV and a second control signal for a camera of the ADV (602), provide the first control signal to the Lidar sensor and the second control signal to the camera (604), and process Lidar output of the Lidar sensor and camera output of the camera to detect one or more features of the Lidar output or camera output (606). In response to detecting the one or more features, the computing device is to adjust the first control signal or the second control signal (608).

Classes IPC  ?

• G05D 13/62 - Commande de la vitesse linéaire; Commande de la vitesse angulaire; Commande de l'accélération ou de la décélération, p.ex. d'une machine motrice caractérisée par l'utilisation de moyens électriques, p.ex. l'emploi de dynamos-tachymétriques, l'emploi de transducteurs convertissant des valeurs électriques en un déplacement
• G01S 17/86 - Combinaisons de systèmes lidar avec des systèmes autres que lidar, radar ou sonar, p.ex. avec des goniomètres

Abrégé

A hardware unit (120) of an ADV (101) comprises an input port to directly receive data from one or more sensors (115) perceiving a driving environment. The hardware unit (120) is coupled with the one or more sensors (115) to perform data processing of the data from one or more sensors (115). The hardware unit (120) comprises a monitor unit (404) to monitor a data rate of output data after the data processing. The hardware unit (120) further comprises a self-adjustment unit (405) to dynamically configure and adjust the data processing based on the data rate of output data. The hardware unit (120) further comprises an output port (412) to transfer the output data to an autonomous driving system (ADS) (110) of the ADV (101) to control the ADV (101) to drive autonomously based on the output data. The hardware unit (120) can improve operation safety of the ADV.

Classes IPC  ?

• G05D 1/00 - Commande de la position, du cap, de l'altitude ou de l'attitude des véhicules terrestres, aquatiques, aériens ou spatiaux, p.ex. pilote automatique
• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• B60W 50/00 - COMMANDE CONJUGUÉE DE PLUSIEURS SOUS-ENSEMBLES D'UN VÉHICULE, DE FONCTION OU DE TYPE DIFFÉRENTS; SYSTÈMES DE COMMANDE SPÉCIALEMENT ADAPTÉS AUX VÉHICULES HYBRIDES; SYSTÈMES D'AIDE À LA CONDUITE DE VÉHICULES ROUTIERS, NON LIÉS À LA COMMANDE D'UN SOUS-ENSEMBLE PARTICULIER - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier

Abrégé

A video system includes a first data gathering node configured to receive a plurality of image streams from a plurality of cameras, respectively. Each of the plurality of cameras captures an environment of an autonomous driving vehicle (ADV). The first data gathering node tags the plurality of image streams with metadata that identifies each of the plurality of image streams, and combines the plurality of image streams with the metadata to form a combined image stream. A second data gathering node is communicatively coupled to the first data gathering node and is to receive the combined image stream from the first data gathering node and output the combined image stream with a second combined image stream.

Classes IPC  ?

• H04N 19/17 - Procédés ou dispositions pour le codage, le décodage, la compression ou la décompression de signaux vidéo numériques utilisant le codage adaptatif caractérisés par l’unité de codage, c. à d. la partie structurelle ou sémantique du signal vidéo étant l’objet ou le sujet du codage adaptatif l’unité étant une zone de l'image, p.ex. un objet

Abrégé

A sensor system for an autonomous driving vehicle (ADV) includes a sensor interface coupled to a plurality of sensors, a host interface coupled to a host system of the ADV, and a self-adaptive sensor transfer unit coupled between the sensor interface and the host interface. The self-adaptive sensor transfer unit includes a sensor monitor module, configured to monitor a data rate of sensor data received from a sensor, and a configuration control module, configured to: receive a target data rate from the host via the host interface; receive the monitored data rate of the sensor data; and control the data rate of the sensor data to be within a threshold of the target data rate.

Classes IPC  ?

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

Abrégé

A method (200, 400, 600) and apparatus (1300, 1400, 1500) for synchronously collecting data, a synchronization determination method (800) and apparatus (1600), a vehicle (110, 1100), and an electronic device (1700), which relate to the field of artificial intelligence, and in particular to the technical field of autonomous driving, computer vision and cloud computing. The specific implementation solution of the method (200) for synchronously collecting data is: in response to receiving a data packet for a predetermined angle (302) from a radar sensor (111, 601), determining first time information (301) of when the radar sensor (111, 601) collects point cloud data at the predetermined angle (302) (S210), wherein the predetermined angle (302) is within an angle-of-view range of an image sensor (112, 602); according to the first time information (301), determining delay information (306) for the image sensor (112, 602) (S220); and sending the delay information (306) to a controller (1000) (S230), such that the controller (1000) controls the image sensor (112, 602) to synchronously collect data with the radar sensor (111, 601) which is rotated to the predetermined angle (302), wherein the data packet comprises the point cloud data collected by the radar sensor (111, 601) at the predetermined angle (302).

Classes IPC  ?

• G01S 17/89 - Systèmes lidar, spécialement adaptés pour des applications spécifiques pour la cartographie ou l'imagerie

Abrégé

A system obtains a performance profile corresponding to times taken to perform an inferencing by a machine learning (ML) model using a different number of processing resources from a plurality of processing resources. The system determines one or more groupings of processing resources from the plurality of processing resources, each grouping includes one or more partitions. The system calculates performance speeds corresponding to each grouping based on the performance profile. The system determines a grouping having a best performance speed from the calculated performance speeds. The system partitions the processing resources based on the determined grouping to perform the inferencing.

Classes IPC  ?

• G06F 9/50 - Allocation de ressources, p.ex. de l'unité centrale de traitement [UCT]

Abrégé

A system determines a timing matrix corresponding to inference times taken for a number of machine learning (ML) models to be executed by a number of processing resources of a computing device. The processing resources includes at least a first and a second type of processing resources. The system applies a service-specific model-first scheduling scheme or a service-specific hardware-first scheduling scheme to obtain corresponding service-specific mappings. The system determines a best mapping from the corresponding service-specific mappings. The system schedules each of the ML models to a corresponding processing resource from the processing resources according to the best mapping. The system executes the ML models using corresponding mapped processing resources.

Classes IPC  ?

• G06F 9/50 - Allocation de ressources, p.ex. de l'unité centrale de traitement [UCT]

Abrégé

Information recommendation system usually involve a multitask problem, which tries to predict not only users' click -through rate (CTR) but also the post-click conversion rate (CVR). At the same time, for multi-functional information systems, there are commonly multiple services for users, such as news feed, search engine, and product suggestions. The prediction/ranking model should be conducted in a multi-scene manner. In the present patent document, embodiments of a unified ranking model for such a multi-task and multi-scene problem are disclosed. The disclosed model explores independent and non-shared embeddings for each task and scene, which reduces the coupling between tasks and scenes. Therefore, new tasks or scenes may be added easily. Besides, a simplified network may be chosen beyond the embedding layer, which largely improves the ranking efficiency for various online services. Extensive offline and online experiments demonstrated the superiority of model embodiments.

Classes IPC  ?

• G06N 3/08 - Méthodes d'apprentissage
• G06Q 30/08 - Ventes aux enchères

Abrégé

Image super-resolution (SR) refers to the process of recovering high-resolution (HR) images from low-resolution (LR) inputs. Blind image SR is a more challenging task which involves unknown blurring kernels and characterizes the degradation process from HR to LR. Embodiments of a variational autoencoder (VAE) are leveraged to train a kernel autoencoder for more accurate degradation representation and more ef-ficient kernel estimation. In one or more embodiments, a kernel-agnostic loss is used to learn more robust kernel features in the latent space from LR inputs without using ground-truth kernel references. In addition, attention-based adaptive pooling is intr-oduced to improve kernel estimation accuracy, and spatially non-uniform kernel fea-tures are passed into SR restoration resulting in additional kernel estimation error to-lerance. Extensive experiments on synthetic and real-world images show that embo-diments of the presented model outperform state-of-the-art methods significantly wi-th the peak signal-to-noise ratio (PSNR) raised considerably.

Classes IPC  ?

• G06T 3/40 - Changement d'échelle d'une image entière ou d'une partie d'image

Abrégé

Modem deep neural network (DNN) models have many layers with a single layer potentially involving large matrix multiplications. Such heavy calculation brings challenges to deploy such DNN models on a single edge device, which has relatively limited computation resources. Therefore, multiple and even heterogeneous edge devices may be required for applications with stringent latency requirements. We provide a model scheduling framework that schedules multiple models on a heterogeneous platform. Multiple-model heterogeneous computing is partitioned into a neural computation optimizer (NCO) part and a neural computation accelerator (NCA) part. The migration, transition, or transformation of DNN models from cloud to edge is handled by the NCO, while the deployment of the transformed DNN models on the heterogeneous platform is handled by the NCA. Such a separation of implementation simplifies task execution and improves the flexibility for the overall framework.

Classes IPC  ?

• G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

Abrégé

Deep neural network (DNN) models have been widely used for user-relevance content prediction. Presented herein is a new user-relevance framework, embodiments of which may be referred as Gating-Enhanced Multi-task Neural Networks (GemNN). In one or more, neural network-based multi-task learning model embodiments herein predict user engagement with content in a coarse-to-fine manner, which gradually reduces content candidates and allows parameter sharing from upstream tasks to downstream tasks to improve the training efficiency. Also, in one or more embodiments, a gating mechanism was introduced between embedding layers and multi-layer perceptions to learn feature interactions and control the information flow fed to MLP layers. Tested embodiments demonstrated considerable improvements over prior approaches.

Classes IPC  ?

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

Abrégé

A driving environment is perceived based on sensor data obtained from a plurality of sensors mounted on the ADV, including detecting a traffic light, where the plurality of sensors includes at least one image sensor. A first sensor setting (402) is applied to the at least one image sensor to capture a first frame, and a second sensor setting (403) is applied to the at least one image sensor to capture a second frame. A color of the traffic light is determined based on sensor data of the at least one image sensor in the first frame. The ADV is controlled to drive autonomously according to the color of the traffic light determined based on sensor data of the at least one image sensor in the first frame and a driving environment perceived based on sensor data of the at least one image sensor in the second frame.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• B60W 40/02 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés aux conditions ambiantes

Abrégé

One application of deep learning method and labelled data is for industrial production or work applications. For such applications implemented with machine learning applications, massive amounts of data are required to train, validate, and/or tune models for better fitting the requirements. However, obtaining such data has typically be costly and difficult. A adaptable process provides data labelling method for work settings. The method take advantage of the work or production processes to label and collect data, which save time and money and improves accuracy. The method prevent or reduce the need for worker training costs and human mistake-triggered data labelling problems. The method also improve data labelling quality and speed-up of the development cycle.

Classes IPC  ?

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

Abrégé

A method facilitates improvement of a deployed neural network model's accuracy without significantly affecting its operation. Online training of the deployed model may be performed using a second neural network model that has higher accuracy than the deployed neural network model. The second neural network model may also be improved online. The method may be deployed in system, such as edge computing environments, in which neural networks deployed at the edge can be centrally monitored and updated.

Classes IPC  ?

• G06N 3/08 - Méthodes d'apprentissage

Abrégé

Network architecture search (NAS) received a lot of attention. The supernet-based differentiable approach is popular because it can effectively share the weights and lead to more efficient search. However, the mismatch between the architecture and weights caused by weight sharing still exists. Moreover, the coupling effects among different operators are also neglected. To alleviate these problems, embodiments of an effective NAS methodology by similarity-based operator ranking are presented herein. With the aim of approximating each layer's output in the supernet, a similarity-based operator ranking based on statistical random comparison is used. In one or more embodiments, then the operator that possibly causes the least change to feature distribution discrepancy is pruned. In one or more embodiments, a fair sampling process may be used to mitigate the operators' Matthew effect that happened frequently in previous supernet approaches.

Classes IPC  ?

• G06N 3/04 - Architecture, p.ex. topologie d'interconnexion
• G06N 3/08 - Méthodes d'apprentissage
• G06F 8/41 - Compilation

Abrégé

Described herein are embodiments of a reinforcement learning based large-scale multi-objective ranking system. Embodiments of the system may be used for optimizing short-video recommendation on a video sharing platform. Multiple competing ranking objective and implicit selection bias in user feedback are the main challenges in real-world platform. In order to address those challenges, multi-gate mixture of experts (MMoE) and soft actor critic (SAC) are integrated together into a MMoE_SAC system. Experiment results demonstrate that embodiments of the MMoE_SAC system may greatly reduce a loss function compared to systems only based on single strategies.

Classes IPC  ?

• G06N 3/04 - Architecture, p.ex. topologie d'interconnexion
• G06N 3/08 - Méthodes d'apprentissage
• G06F 16/70 - Recherche d’informations; Structures de bases de données à cet effet; Structures de systèmes de fichiers à cet effet de données vidéo
• G06F 16/957 - Optimisation de la navigation, p.ex. mise en cache ou distillation de contenus

Abrégé

Concept graphs are created as universal taxonomies for text understanding in the open domain knowledge. The nodes in concept graphs include both entities and concepts. The edges are from entities to concepts, showing that an entity is an instance of a concept. Presented herein are embodiments that handle the task of learning interpretable relationships from open domain facts to enrich and refine concept graphs. In one or more embodiments, the Bayesian network structures are learned from open domain facts as the interpretable relationships between relations of facts and concepts of entities. Extensive experiments were conducted on English and Chinese datasets. Compared to the state-of-the-art methods, the learned network structures improve the identification of concepts for entities based on the relations of entities on both English and Chinese datasets.

Classes IPC  ?

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

Abrégé

Presented herein are novel embodiments for converting a given speech audio or text into a photo-realistic speaking video of a person with synchronized, realistic, and expressive body dynamics. 3D skeleton movements are generated from the audio sequence using a recurrent neural network, and an output video is synthesized via a conditional generative adversarial network. To make movements realistic and expressive, the knowledge of an articulated 3D human skeleton and a learned dictionary of personal speech iconic gestures may be embedded into the generation process in both learning and testing pipelines. The former prevents the generation of unreasonable body distortion, while the later helps the model quickly learn meaningful body movement with a few videos. To produce photo-realistic and high-resolution video with motion details, a part-attention mechanism is inserted in the conditional GAN, where each detailed part is automatically zoomed in to have their own discriminators.

Classes IPC  ?

• G06T 19/00 - Transformation de modèles ou d'images tridimensionnels [3D] pour infographie

Abrégé

Methods and systems for reliably detecting malfunctions in a variety of software or hardware components (405, 407) in an autonomous driving vehicle (ADV) (101), a redundant system (327) can be provided on an independent computing device in an ADV (101) to check for malfunctions in a number of software or hardware components (405, 407). When no malfunction occurs in the ADV (101), an autonomous driving system (ADS) (110) in the ADV (101) operates to drive the ADV (101), while the redundant system (327) can monitor the ADS (110) in a standby mode. In the event of a malfunction, the redundant system (327) can take over the control of the ADV (101), and take appropriate actions based on a severity level of the malfunction.

Classes IPC  ?

• G05B 9/03 - Dispositions de sécurité électriques avec une boucle à canal multiple, c. à d. systèmes de commande redondants

Abrégé

Systems and methods of video inpainting for autonomous driving are disclosed. For example, the method stitches a multiplicity of depth frames into a 3D map, where one or more objects in the depth frames have previously been removed. The method further projects the 3D map onto a first image frame to generate a corresponding depth map, where the first image frame includes a target inpainting region. For each target pixel within the target inpainting region of the first image frame, based on the corresponding depth map, the method further maps the target pixel within the target inpainting region of the first image frame to a candidate pixel in a second image frame. The method further determines a candidate color to fill the target pixel. The method further performs Poisson image editing on the first image frame to achieve color consistency at a boundary and between inside and outside of the target inpainting region of the first image frame. For each pixel in the target inpainting region of the first image frame, the method further traces the pixel into neighboring frames and replacing an original color of the pixel with an average of colors sampled from the neighboring frames.

Classes IPC  ?

• H04N 13/00 - Systèmes vidéo stéréoscopiques; Systèmes vidéo multi-vues; Leurs détails
• G06T 7/50 - Récupération de la profondeur ou de la forme

Abrégé

A method, apparatus, and system for estimating a moving speed of a detected pedestrian at an autonomous driving vehicle (ADV) is disclosed. A pedestrian is detected in a plurality of frames of point clouds generated by a LIDAR device installed at an autonomous driving vehicle (ADV). In each of at least two of the plurality of frames of point clouds, a minimum bounding box enclosing points corresponding to the pedestrian excluding points corresponding to limbs of the pedestrian is generated. A moving speed of the pedestrian is estimated based at least in part on the minimum bounding boxes across the at least two of the plurality of frames of point clouds. A trajectory for the ADV is planned based at least on the moving speed of the pedestrian. Thereafter, control signals are generated to drive the ADV based on the planned trajectory.

Classes IPC  ?

• G01S 17/93 - Systèmes lidar, spécialement adaptés pour des applications spécifiques pour prévenir les collisions

Abrégé

A method, apparatus, and system for determining whether all extrinsic matrices are accurate is disclosed. A plurality of post-LIDAR fusion point clouds that are based on simultaneous outputs from a plurality of LIDAR devices installed at one ADV are obtained (610). The obtained plurality of point clouds are filtered to obtain a first set of points comprising all points in the plurality of point clouds that fall within a region of interest (620). Each point in the first set of points corresponds to one coordinate value on the axis in the up-down direction (630). A distribution of the first plurality of coordinate values is obtained (640). A quantity of peaks in the distribution of the first plurality of coordinate values is determined (650). Whether all extrinsic matrices associated with the plurality of LIDAR devices are accurate is determined based on the quantity of peaks in the distribution of the first plurality of coordinate values (660).

Classes IPC  ?

• G06T 7/30 - Détermination des paramètres de transformation pour l'alignement des images, c. à d. recalage des images

Abrégé

An inertial navigation system (530), including a global navigation satellite system (GNSS) receiver unit (532), a first inertial measurement unit (IMU) (534A) and a second IMU (534B). The system (530) may further include a first micro-controller unit (MCU) (536A) coupled to the first IMU (534A) and the GNSS receiver unit (532) to receive data from the first IMU (534A) and the GNSS receiver unit (532) and a second MCU (536B) coupled to the second IMU (534B) and the GNSS receiver unit (532) to receive data from the second IMU (534B) and the GNSS receiver unit(532).

Classes IPC  ?

• G01C 21/00 - Navigation; Instruments de navigation non prévus dans les groupes
• G01C 21/16 - Navigation; Instruments de navigation non prévus dans les groupes en utilisant des mesures de la vitesse ou de l'accélération exécutées à bord de l'objet navigant; Navigation à l'estime en intégrant l'accélération ou la vitesse, c. à d. navigation par inertie
• G01S 19/49 - Détermination de position en combinant ou en commutant entre les solutions de position dérivées du système de positionnement par satellite à radiophares et les solutions de position dérivées d'un autre système l'autre système étant un système de position inertielle, p.ex. en hybridation lâche

Abrégé

A pull over method based on quadratic programming for path planning, in response to a request to pull over an ADV at a destination point (720) at a side of a lane (605,705), a path including a first segment (721), a second segment (722) and a transition point (606,706) is planned. The transition point (606,706) is determined based on at least one of a distance to the destination point (720) or a predetermined distance to a boundary (603,703,604,704) of the side of the lane (605,705). The first segment (721) from a start point to the transition point (606,706) is generated by using a quadratic programming (QP) operation. The second segment (722) from the transition point (606,706) to the destination is generated based on a shape of the boundary (603,703,604,704). The ADV is controlled to pull over to the destination point (720) according to the planned path.

Classes IPC  ?

• B60W 40/02 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés aux conditions ambiantes
• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire

Abrégé

In response to a request to park an ADV (101, 300, 603, 803) in a parking lot (601, 801), a set of parking-trajectories (621, 622, 623, 624, 625, 626, 821) associated with a set of predetermined locations (611, 612, 613, 614, 615, 616, 811) near one or more parking spots (602, 802) in the parking lot (601, 801) may be obtained, where the set of parking-trajectories (621, 622, 623, 624, 625, 626, 821) was previously generated based on prior collected planning and control data of the parking lot (601, 801). Each parking-trajectory of the set of parking-trajectories (621, 622, 623, 624, 625, 626, 821) may correspond to one parking spot of the one or more parking spots(602, 802) in the parking lot (601, 801). A parking-trajectory may be selected from the set of parking-trajectories (621, 622, 623, 624, 625, 626, 821) based on a current location (810) of the ADV (101,300,603,803). The ADV (101, 300, 603, 803) may be controlled to park in a corresponding parking spot according to the selected parking-trajectory.

Classes IPC  ?

• B60W 30/06 - Manœuvre automatique de stationnement
• G08G 1/14 - Systèmes de commande du trafic pour véhicules routiers indiquant des places libres individuelles dans des parcs de stationnement

Abrégé

A method, apparatus, and system for filtering a point cloud generated by a LIDAR device in an autonomous vehicle is disclosed. A point cloud comprising a plurality of points is generated based on outputs of the LIDAR device. The point cloud is filtered to remove a first set of points in the plurality of points that correspond to noise based on one or more of: point intensity measurements, distances between points, or a combination thereof. Perception data is generated based on the filtered point cloud. Operations of the autonomous vehicle are controlled based on the perception data.

Classes IPC  ?

• G01S 17/93 - Systèmes lidar, spécialement adaptés pour des applications spécifiques pour prévenir les collisions

Abrégé

A method, apparatus, and system for determining a low-height obstacle based on outputs of a LIDAR device in an autonomous vehicle is disclosed. A point cloud comprising a plurality of points is generated based on outputs of a LIDAR device. For each point within a first number of lowest rings of points, a neighboring point in a same ring to a first direction is determined, and a first and a second coordinate values-related differences are determined. A first, a second, a third, and a fourth quantities are determined based on the first and second differences. In response to determining that the first, the second, the third, and the fourth quantities satisfy a predetermined condition, a low-height obstacle is determined based on the points within the first number of lowest rings of points. Operations of an autonomous vehicle are controlled based at least in part on the determined low-height obstacle.

Classes IPC  ?

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

Abrégé

A method for synchronizing sensor data of an autonomous driving vehicle includes determining, by a processing device of an inertial navigation system (INS), that global navigation satellite system (GNSS) data is unavailable (S602) and identifying an alternative source of time information (S604). The method further includes retrieving time information from the alternative source (S606) and synchronizing sensor data with the time information from the alternative source of time information (S608).

Classes IPC  ?

• G01C 21/16 - Navigation; Instruments de navigation non prévus dans les groupes en utilisant des mesures de la vitesse ou de l'accélération exécutées à bord de l'objet navigant; Navigation à l'estime en intégrant l'accélération ou la vitesse, c. à d. navigation par inertie

Abrégé

Disclosed herein is a computer-implemented method for operating an autonomous driving vehicle (ADV), comprising: determining a starting point, a set of one or more way points, and a destination point of a first route along which the ADV is to be driven(1001);determining all lane segments near the starting point, the set of way points, and the destination point within a predetermined threshold distance respectively(1002); determining a set of route candidates using an A-star (A*) searching algorithm based on a set of nodes representing all lane segments near the starting point, the set of way points, and the destination point respectively(1003); selecting a second route from the set of route candidates based on respective costs of the set of route candidates(1004); and controlling the ADV to drive along the selected route autonomously(1005).

Classes IPC  ?

• G01C 21/32 - Structuration ou formatage de données cartographiques
• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire

Abrégé

A method, apparatus, and system for filtering obstacle candidates determined based on outputs of a LIDAR device in an autonomous vehicle is disclosed. A point cloud comprising a plurality of points is generated based on outputs of the LIDAR device (610). One or more obstacle candidates are determined based on the point cloud (620). The one or more obstacle candidates are filtered to remove a first set of obstacle candidates in the one or more obstacle candidates that correspond to noise based at least in part on characteristics associated with points that correspond to each of the one or more obstacle candidates (630). One or more recognized obstacles comprising the obstacle candidates that have not been removed are determined (640). Operations of an autonomous vehicle are controlled based on the recognized obstacles (650).

Classes IPC  ?

• G06T 5/00 - Amélioration ou restauration d'image
• G01S 17/93 - Systèmes lidar, spécialement adaptés pour des applications spécifiques pour prévenir les collisions

Abrégé

In one embodiment, an exemplary method of autonomously charging an autonomous driving vehicle includes receiving, from a sensor in an autonomous driving vehicle (ADV), indication that a batter level of the ADV falls below a threshold; and selecting a charging pile from a plurality of charging piles on a high definition map based on information received from a cloud server. The method further includes generating a first trajectory based on a current location of the ADV and a location of the selected charging pile, the first trajectory connecting a first point representing the current location of the ADV to a second point at the selected charging pile, and including a first segment and a second segment. The method further includes driving forward along the first segment of the first trajectory, and driving backward along the second segment of the first trajectory when the ADV drives towards the selected charging pile along the first trajectory.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• H02J 7/00 - Circuits pour la charge ou la dépolarisation des batteries ou pour alimenter des charges par des batteries

Abrégé

A method for calculating a location of an autonomous driving vehicle (ADV) includes: receiving new global navigation satellite system (GNSS) data (602), identifying a first previously estimated location from a plurality of previously estimated locations with a timestamp that is closest to the timestamp of the new GNSS data (604), identifying a second previously estimated location from the plurality of previously estimated locations with a most recent timestamp (606), calculating a difference between the first previously estimated location and the second previously estimated location (608), adjusting the new GNSS data based on the difference (610), and calculating a current estimated location of the ADV based on the adjusted GNSS data (612).

Classes IPC  ?

• G01S 19/47 - Détermination de position en combinant les mesures des signaux provenant du système de positionnement satellitaire à radiophares avec une mesure supplémentaire la mesure supplémentaire étant une mesure inertielle, p.ex. en hybridation serrée
• G01S 19/42 - Détermination de position
• G01C 21/16 - Navigation; Instruments de navigation non prévus dans les groupes en utilisant des mesures de la vitesse ou de l'accélération exécutées à bord de l'objet navigant; Navigation à l'estime en intégrant l'accélération ou la vitesse, c. à d. navigation par inertie

Abrégé

A computer-implement method (1200) for operating an ADV (101, 810) is disclosed, the method(1200) comprising: determining a starting point (803a, 803b, 803c, 803d, 903) and an ending point (804a, 804b, 804c, 804d, 904) of a route along which the ADV (101, 810) is to be driven (1201); determining whether each of the starting point (803a, 803b, 803c, 803d, 903) and the ending point (804a, 804b, 804c, 804d, 904) is within a first driving area of a first type having a lane boundary or a second driving area of a second type as an open space that is without a lane boundary (1202); dividing the route into a first route segment and a second route segment based on the determining whether each of the starting point (803a, 803b, 803c, 803d, 903) and the ending point (804a, 804b, 804c, 804d, 904) is within the first driving area or the second driving area (1203); operating in one of an on-lane mode or an open-space mode to plan a first trajectory for the first route segment and operating in one of the on-lane mode or the open-space mode to plan a second trajectory for the second route segment (1204).

Classes IPC  ?

• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire
• G01C 21/26 - Navigation; Instruments de navigation non prévus dans les groupes spécialement adaptés pour la navigation dans un réseau routier
• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions

Abrégé

A computer-implement method for operating an ADV is disclosed. A first trajectory for the ADV to drive along is planned (S1401). The ADV is to autonomously drive along the first trajectory (S1402). An obstacle in an affected region of the ADV is detected based on sensor data obtained from a plurality of sensors mounted on the ADV (S1403). An expected residence time of the obstacle in the affected region is determined (S1404). Whether to plan a second trajectory or to wait for the obstacle to leave the affected region is determined based on the expected residence time of the obstacle in the affected region (S1405). A second trajectory for the ADV to drive along is planned and the ADV is to autonomously drive along the second trajectory, or the ADV is to wait for the obstacle to leave the affected region and to autonomously drive along the first trajectory afterwards (S1406).

Classes IPC  ?

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

Abrégé

Systems and methods are disclosed for collecting driving data from simulated autonomous driving vehicle (ADV) driving sessions and real-world ADV driving sessions. The driving data is processed to exclude manual (human) driving data and to exclude data corresponding to the ADV being stationary (not driving). Data can further be filtered based on driving direction: forward or reverse driving. Driving data records are time stamped. The driving data can be aligned according to the timestamp, and then a standardized set of metrics is generated from the collected, filtered, and time-aligned data. The standardized set of metrics are used to grade the performance the control system of the ADV, and to generate an updated ADV controller, based on the standardized set of metrics. The methods provide a systematic, comprehensive and automatic test tool for autonomous vehicle control system.

Classes IPC  ?

• B60W 30/18 - Propulsion du véhicule
• B60W 30/00 - Fonctions des systèmes d'aide à la conduite des véhicules routiers non liées à la commande d'un sous-ensemble particulier, p.ex. de systèmes comportant la commande conjuguée de plusieurs sous-ensembles du véhicule

Abrégé

A lateral control error of a steering system of an ADV is determined, which includes iteratively performing following operations for a predetermined time period. Whether the ADV is moving within a predetermined proximity of a current moving direction is determined. Next, whether a road condition of a road on which the ADV is driving satisfies a predetermined road condition is determined. Then, a first steering feedback of the ADV in response to a prior steering control command is measured. Thereafter, the lateral control error is determined based on at least a portion of the first steering feedback over the predetermined time period. Further, a steering command in view of the lateral control error of the steering system is generated. Finally, the steering command is applied to control the ADV to compensate the lateral control error of the steering system.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• B60W 30/02 - Commande de la stabilité dynamique du véhicule

Abrégé

A deep residual network dedicated to color filter array mosaic pattern is disclosed. A mosaic stride convolution layer is introduced to match the mosaic pattern of a multispectral filter arrays (MSFA) or a color filter array raw image. A data augmentation using MSFA shifting and dynamic noise are applied to make the model robust to different noise levels. Network optimization criteria may be created by using the noise standard deviation to normalize the L 1loss function. Comprehensive experiments demonstrates that the disclosed deep residual network outperform the state-of-the-art denoising algorithms in MSFA field.

Classes IPC  ?

• G06T 5/00 - Amélioration ou restauration d'image

Abrégé

In one embodiment, an exemplary method includes the operations of receiving a raw reference line representing a route from a first location to a second location associated with an autonomous driving vehicle (ADV); and smoothing the raw reference line using a Quadratic programming (QP) spline smoother to generate a smoothed reference line. The method further includes the operations of identifying one or more segments on the smoothed reference line, each of the identified reference line segments including a curvature that exceeds a predetermined size; and smoothing each of the one or more identified reference line segments using a spiral smoother, including optimizing each identified curvature in view of a set of constraints, such that an output of the objective function reaches a minimum value while the set of constraints are satisfied; and controlling the ADV using the smoothed reference line.

Classes IPC  ?

• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire

Abrégé

A simulation method of an autonomous driving vehicle (ADV) includes capturing first data that includes a control command output by an autonomous vehicle controller of the ADV, and capturing second data that includes the control command being implemented at a control unit of the ADV. The control command, for example, a steering command, a braking command, or a throttle command, is implemented by the ADV to affect movement of the ADV. A latency model is determined based on comparing the first data with the second data, where the latency model defines time delay and/or amplitude difference between the first data and the second data. The latency model is applied in a virtual driving environment.

Classes IPC  ?

• G06F 9/445 - Chargement ou démarrage de programme
• G06F 17/00 - TRAITEMENT ÉLECTRIQUE DE DONNÉES NUMÉRIQUES Équipement ou méthodes de traitement de données ou de calcul numérique, spécialement adaptés à des fonctions spécifiques
• G05D 1/00 - Commande de la position, du cap, de l'altitude ou de l'attitude des véhicules terrestres, aquatiques, aériens ou spatiaux, p.ex. pilote automatique
• B60W 10/20 - Commande conjuguée de sous-ensembles de véhicule, de fonction ou de type différents comprenant la commande des systèmes de direction
• B60W 10/06 - Commande conjuguée de sous-ensembles de véhicule, de fonction ou de type différents comprenant la commande des ensembles de propulsion comprenant la commande des moteurs à combustion
• B60W 10/18 - Commande conjuguée de sous-ensembles de véhicule, de fonction ou de type différents comprenant la commande des systèmes de freinage

Abrégé

A dynamic cost function can be used to collect real-time values of a set of parameters, and use the real-time values to constantly adjust a preferred safety distance where the ADV can be stopped ahead of the obstacle. A method includes determining a first distance to the obstacle in response to detecting an obstacle ahead of the ADV; and for each of a number of iterations, collecting a real-time value for each of a set of parameters, determining an offset to the first distance using the real-time value for each of the set of parameters, calculating a second distance based on the first distance and the offset, and controlling the ADV in view of the second distance using an expected value of each of the set of parameters, such that the ADV can stop at a point having the second distance to the obstacle.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• B60W 30/09 - Entreprenant une action automatiquement pour éviter la collision, p.ex. en freinant ou tournant

Abrégé

Tensor decomposition can be advantageous for compressing deep neural networks (DNNs). In many applications of DNNs, reducing the number of parameters and computation workload is helpful to accelerate inference speed in deployment. Modern DNNs comprise multiple layers with multi-array weights where tensor decomposition is a natural way to perform compression-in which the weight tensors in convolutional layers or fully-connected layers are decomposed with specified tensor ranks (e.g., canonical ranks, tensor train ranks). Conventional tensor decomposition with DNNs involves selecting ranks manually, which requires tedious human efforts to finetune the performance. Accordingly, presented herein are rank selection embodiments, which are inspired by reinforcement learning, to automatically select ranks in tensor decomposition. Experimental results validate that the learning-based rank selection embodiments significantly outperform hand-crafted rank selection heuristics on a number of tested datasets, for the purpose of effectively compressing deep neural networks while maintaining comparable accuracy.

Classes IPC  ?

• G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

Abrégé

A data processing system includes a central processing unit (CPU) (107,109) and ac-celerator cards coupled to the CPU (107,109) over a bus, each of the accelerator card-s having a plurality of data processing (DP) accelerators to receive DP tasks from the CPU (107,109) and to perform the received DP tasks. At least two of the accelerator cards are coupled to each other via an inter-card connection, and at least two of the DP accelerators are coupled to each other via an inter-chip connection. Each of the inter-card connection and the inter-chip connection is capable of being dynamically activated or deactivated, such that in response to a request received from the CPU (107,109), any one of the accelerator cards or any one of the DP accelerators within any one of the accelerator cards can be enabled or disabled to process any one of the DP tasks received from the CPU (107,109).

Classes IPC  ?

• G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

Abrégé

According to various embodiments, methods and systems are provided to accelerate artificial intelligence (AI) model training with advanced interconnect communication technologies and systematic zero-value compression over a distributed training system. During each iteration of a Scatter-Reduce process performed on a cluster of processors (117,119,121) arranged in a logical ring to train a neural network model (701), a processor receives a compressed data block from a prior processor (107) in the logical ring, performs an operation on the received compressed data block and a compressed data block generated on the processor (117,119,121) to obtain a calculated data block, and sends the calculated data block to a following processor (117,119,121) in the logical ring (702). A compressed data block calculated from corresponding data blocks from the processors (117,119,121) can be identified on each processor (117,119,121) and distributed to each other processor (117,119,121) and decompressed therein for use in the AI model training (703).

Classes IPC  ?

• G06T 1/20 - Architectures de processeurs; Configuration de processeurs p.ex. configuration en pipeline
• G06N 20/00 - Apprentissage automatique

Abrégé

Deep neural networks (DNN) model quantization may be used to reduce storage and computation burdens by decreasing the bit width. A method for cursor-based adaptive quantization for a neural network, a multiple bits quantization mechanism is formulated as a differentiable architecture search (DAS) process with a continuous cursor that represents a possible quantization bit. The cursor-based DAS adaptively searches for a quantization bit for each layer. The DAS process may be accelerated via an alternative approximate optimization process, which is designed for mixed quantization scheme of a DNN model. A new loss function is used in the search process to simultaneously optimize accuracy and parameter size of the model. In a quantization step, the closest two integers to the cursor may be adopted as the bits to quantize the DNN together to reduce the quantization noise and avoid the local convergence problem.

Classes IPC  ?

• G06N 3/08 - Méthodes d'apprentissage
• G06N 3/04 - Architecture, p.ex. topologie d'interconnexion

Abrégé

A method is provided. The method includes determining a first set of data acquisition characteristics of a first sensor of an autonomous driving vehicle. The method also includes determining a second set of data acquisition characteristics of a second sensor of the autonomous driving vehicle. The method further includes synchronizing a first data acquisition time of the first sensor and a second data acquisition time of the second sensor, based on the first set of data acquisition characteristics and the second set of data acquisition characteristics. The first sensor obtains first sensor data at the first data acquisition time. The second sensor obtains second sensor data at the second data acquisition time.

Classes IPC  ?

• G01D 21/02 - Mesure de plusieurs variables par des moyens non couverts par une seule autre sous-classe

Abrégé

A method of verifying operation of a sensor, includes causing a sensor to obtain sensor data (1215) at a first time, wherein the sensor obtains the sensor data by emitting waves towards a detector; determining that the detector has detected the waves (1220) at a second time; receiving the sensor data from the sensor (1225) at a third time; and verifying operation of the sensor (1230) based on at least one of the first time, the second time, or the third time.

Classes IPC  ?

• H04J 3/06 - Dispositions de synchronisation

Abrégé

A computer-implemented method of analyzing sensor synchronization in an autonomous driving vehicle (ADV), comprising: acquiring raw sensor data from a first sensor and a second sensor mounted on the ADV, the raw sensor data describing a target object in a surrounding environment of the ADV; and generating an accuracy map based on the raw sensor data in view of timestamps extracted from the raw sensor data. The method further include the operations of generating a first bounding box and a second bounding box around the target object using the raw sensor data; and performing an analysis of the first and second bounding boxes and the accuracy map using a predetermined algorithm in view of one or more pre-configured sensor settings to determine whether the first sensor and the second sensor are synchronized with each other.

Classes IPC  ?

• G01S 7/497 - Moyens de contrôle ou de calibrage

Abrégé

Described herein are embodiments for end-to-end reinforcement learning based coreference resolution models to directly optimize coreference evaluation metrics. Embodiments of a reinforced policy gradient model are disclosed to incorporate reward associated with a sequence of coreference linking actions. Furthermore, maximum entropy regularization may be used for adequate exploration to prevent a model embodiment from prematurely converging to a bad local optimum. Experiments on datasets compared with state-of-the-art methods verified the effectiveness of embodiments.

Classes IPC  ?

• G06F 16/30 - Recherche d’informations; Structures de bases de données à cet effet; Structures de systèmes de fichiers à cet effet de données textuelles non structurées

Abrégé

Systems and methods for multispectral image demosaicking using deep panchromatic image guided residual interpolation are disclosed. ResNet-based deep learning models are disclosed to reconstruct the full-resolution panchromatic image from multispectral filter array (MSFA) mosaic image. Reconstructed deep panchromatic image (DPI) is deployed as the guide to recover the full-resolution multispectral image using a two-pass guided residual interpolation methodology. Experiment results demonstrate that the disclosed methods outperform some state-of-the-art conventional and deep learning demosaicking methods both qualitatively and quantitatively.

Classes IPC  ?

• G06T 5/00 - Amélioration ou restauration d'image

Abrégé

A method to perform video compression for ADV is disclosed. The method receives multiple frames of image data from multiple cameras. Metadata are appended to each frame of the image data to generate one of multiple frames of uncompressed image data as the image data are received. The frames of uncompressed image data may be stored. To compress the image data later, the method retrieves the frames of uncompressed image data, extracts the metadata from each frame of the uncompressed image data to generate one of multiple frames of processed image data. The method compresses each frame of the processed image data with the metadata extracted to generate one of multiple frames of compressed image data. The method reattaches the metadata to a corresponding frame of the compressed image data to generate one of multiple compressed image frames. The metadata supports time synchronization and error handling of the image data.

Classes IPC  ?

• H04N 19/17 - Procédés ou dispositions pour le codage, le décodage, la compression ou la décompression de signaux vidéo numériques utilisant le codage adaptatif caractérisés par l’unité de codage, c. à d. la partie structurelle ou sémantique du signal vidéo étant l’objet ou le sujet du codage adaptatif l’unité étant une zone de l'image, p.ex. un objet

Abrégé

An ADV includes a method to combine data from multiple sensors. The method compresses video data from a camera to generate compressed video data. The compressed video data are segmented. The method time synchronizes each segment of the compressed video data with data from other sensors. The method then combines each segment of the compressed video data with the corresponding time-synchronized sensor data for the other sensors. In one embodiment, each segment of the compressed video data is independently decodable. In another embodiment, each segment of the compressed video data includes a compressed video unit that is prepended with a buffered portion of the compressed video data that immediately precede the compressed video unit. The length of the compressed video unit is smaller than the length of the independently decodable segment to offer finer granularity in time synchronizing the compressed video data with the other sensor data with a tradeoff.

Classes IPC  ?

• H04L 29/08 - Procédure de commande de la transmission, p.ex. procédure de commande du niveau de la liaison

Abrégé

Based on sensor data obtained from a variety of sensors, a driving environment surrounding an autonomous driving vehicle (ADV) is perceived, including perceiving and identifying one or more obstacles. A trajectory is planned based on the perception data according to a set of rules to drive the ADV navigating through the driving environment. Trajectory data representing the trajectory is generated, where the trajectory data includes information indicating target or expected vehicle states at different points in time along the trajectory. The trajectory data is then transmitted in a sequence or a stream of one or more controller area network (CAN) messages to an electronic control unit (ECU) of the ADV over a CAN bus. The ECU is configured to generate and issue one or more control commands (e.g., throttle, brake, steering commands) based on the trajectory data to drive the ADV according to the trajectory.

Classes IPC  ?

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

Abrégé

A method for generating a reference line for operating an autonomous driving vehicle, includes determining a first ending reference point having a smallest curvature among a plurality of points within a first defined distance along a path (602), generating a first reference line based on a first initial reference point and the first ending reference point (604), determining a second ending reference point having a smallest curvature among a plurality of points within a second defined distance along the path (606), generating a second reference line based on the first and second ending reference points and an end section of the first reference line (608), connecting the first and second reference lines (610), and controlling the autonomous driving vehicle along the connected first reference line and the second reference line (612).

Classes IPC  ?

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

Abrégé

Provided is a method for operating an autonomous driving vehicle (ADV). The method includes the operation of segmenting a raw reference line into a plurality of reference line segments, including a first reference line segment, a second reference line segment, and a third reference line segment in a sequential order, in response to receiving the raw reference line representing a route from a first location to a second location associated with the ADV. The method further includes the operations of smoothing the first reference line segment and the third reference line segment using a Quadratic programming (QP) spline smoother; and smoothing the second reference line segment using a spiral smoother. Smoothed reference line segments from the plurality of reference line segments are connected to generate a smoothed reference line, which is to be used as a reference line of the route to control the ADV.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• B60W 30/02 - Commande de la stabilité dynamique du véhicule

Abrégé

A method for monitoring safety of an autonomous driving vehicle (ADV) includes: receiving, by a vehicle controller, one or more error message from a patrol module, the one or more error messages generated by an autonomous driving system of the ADV operating in an autonomous mode, the patrol module monitoring the autonomous driving system (301); evaluating a status of the autonomous driving system based on the one or more error messages (302); and keeping the ADV in the autonomous mode or switching it to a manual mode based on the status of the autonomous driving system (303).

Classes IPC  ?

• B60W 50/02 - COMMANDE CONJUGUÉE DE PLUSIEURS SOUS-ENSEMBLES D'UN VÉHICULE, DE FONCTION OU DE TYPE DIFFÉRENTS; SYSTÈMES DE COMMANDE SPÉCIALEMENT ADAPTÉS AUX VÉHICULES HYBRIDES; SYSTÈMES D'AIDE À LA CONDUITE DE VÉHICULES ROUTIERS, NON LIÉS À LA COMMANDE D'UN SOUS-ENSEMBLE PARTICULIER - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier pour préserver la sécurité en cas de défaillance du système d'aide à la conduite, p.ex. en diagnostiquant ou en palliant à un dysfonctionnement
• B60W 30/182 - Sélection entre plusieurs modes opératoires, p.ex. confort ou sportif

Abrégé

A computer-implemented method, apparatus, and system (1500) for receiving and calibrating lateral deviation values and for controlling an autonomous vehicle to correct the lateral deviation are provided. In every perception and planning cycle, a single lateral deviation value representative of an estimated autonomous vehicle lateral deviation from a reference line is generated based on camera detection. The deviation value for a present cycle is received. A calibrated deviation value can be updated for the present cycle based on the received deviation value and a Gaussian distribution model. Control signals for the present cycle are generated to drive the autonomous vehicle to at least partially correct the autonomous vehicle lateral deviation based on the updated calibrated deviation value.

Classes IPC  ?

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

Abrégé

A computer-implemented method for discretizing lane markings and for generating a lane reference line is disclosed. A polynomial defined over an (x, y) coordinate system is received, the polynomial being representative of at least a portion of a lane boundary line. A length of the polynomial is determined. The polynomial is discretized, comprising determining a plurality of discretization points on the polynomial to represent the polynomial, wherein a first discretization piint is a first end of the polynomial, wherein subsequent discretization points are determined successively until the polynomial is completely discretized, and wherein each discretization point other than the first discretization point is determined based at least in part on a slope of the polynomial at a previous discretization point. Thereafter, a lane reference line comprising a plurality of points is generated based on the discretized polynomial. A non-transitory machine-readable medium having instructions stored therein and a data processing system for discretizing lane markings and for generating a lane reference line are also disclosed.

Classes IPC  ?

• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire
• B60W 30/14 - Régulateur d'allure
• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions

Abrégé

A system receives a stream of frames of point clouds from one or more LIDAR sensors of an ADV and corresponding poses in real-time (1401). The system extracts segment information for each frame of the stream based on geometric or spatial attributes of points in the frame, where the segment information includes one or more segments of at least a first frame corresponding to a first pose (1402). The system registers the stream of frames based on the segment information (1403). The system generates a first point cloud map for the stream of frames based on the frame registration (1404).

Classes IPC  ?

• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire

Abrégé

In one embodiment, a system generates an occupancy grid map based on an initial frame of point clouds. The system receives one or more subsequent frames of the point clouds. For each of the subsequent frames, the system updates an occupancy grid map based on the subsequent frame. The system identifies one or more problematic voxels based on the update, the system determines whether the problematic voxels belong to a wall object, and in response to determining that the problematic voxel s belong to a wall object, the system flags the problematic voxels as ghost effect voxels for the subsequent frame.

Classes IPC  ?

• G06T 15/00 - Rendu d'images tridimensionnelles [3D]

Abrégé

A method for extracting point cloud features for use in localizing an autonomous driving vehicle (ADV) includes selecting a first set of keypoints from an online point cloud, the online point cloud generated by a LiDAR device on the ADV for a predicted pose of the ADV (701); and extracting a first set of feature descriptors from the first set of keypoints using a feature learning neural network running on the ADV (703). The method further includes locating a second set of keypoints on a pre-built point cloud map, each keypoint of the second set of keypoints corresponding to a keypoint of the first set of keypoint (705); extracting a second set of feature descriptors from the pre-built point cloud map (707); and estimating a position and orientation of the ADV based on the first set of feature descriptors, the second set of feature descriptors, and a predicted pose of the ADV.

Classes IPC  ?

• G01S 7/40 - Moyens de contrôle ou d'étalonnage
• G01C 21/00 - Navigation; Instruments de navigation non prévus dans les groupes

Abrégé

A method for solution inference using neural networks in LiDAR localization includes constructing a cost volume in a solution space for a predicted pose of an autonomous driving vehicle (ADV). The cost volume includes a number of sub volumes. Each sub volume represents a matching cost between a keypoint from an online point cloud and a corresponding keypoint on a pre-built point cloud map (1001). The method further includes regularizing the cost volume using convention neural networks (CNNs) to refine the matching costs (1003); and inferring, from the regularized cost volume, an optimal offset of the predicted pose. The optimal offset can be used to determine a location of the ADV (1005).

Classes IPC  ?

• G01S 17/89 - Systèmes lidar, spécialement adaptés pour des applications spécifiques pour la cartographie ou l'imagerie
• G06T 7/73 - Détermination de la position ou de l'orientation des objets ou des caméras utilisant des procédés basés sur les caractéristiques

Abrégé

A method for temporal smoothness in localization results for an autonomous driving vehicle includes creating a probability offset volume that represents an overall matching cost between a first set of keypoints from the online point cloud and a second set of keypoints from a pre-built point cloud map for each of a series of sequential light detection and ranging (LiDAR) frames in an online point cloud (1301). The method further includes compressing the probability offset volume into multiple probability vectors across a X dimension, a Y dimension and a yaw dimension (1303); providing each probability vector of the probability offset volume to a number of recurrent neural networks (RNNs) (1305); and generating, by the RNNs, a trajectory of location results across the plurality of sequential LiDAR frames (1307).

Classes IPC  ?

• G01S 7/40 - Moyens de contrôle ou d'étalonnage
• G01C 21/00 - Navigation; Instruments de navigation non prévus dans les groupes

Abrégé

A system for registration of point clouds for autonomous driving vehicles is provided. The system receives a number of point clouds and corresponding poses from the autonomous driving vehicles equipped with LIDAR sensors capturing point clouds of a navigable area to be mapped, where the point clouds correspond to a first coordinate system. The system partitions the point clouds and the corresponding poses into one or more loop partitions based on navigable loop information captured by the point clouds. For each of the loop partitions, the system applies an optimization model to point clouds corresponding to the loop partition to register the point clouds. The system merges the one or more loop partitions together using a pose graph algorithm, where the merged partitions of point clouds are utilized to perceive a driving environment surrounding the autonomous driving vehicles.

Classes IPC  ?

• G01C 21/32 - Structuration ou formatage de données cartographiques
• G01C 21/26 - Navigation; Instruments de navigation non prévus dans les groupes spécialement adaptés pour la navigation dans un réseau routier

Abrégé

In one embodiment, a system receives a number of point clouds captured by one or more LIDAR sensors of an ADV and corresponding poses(1801). The system receives a number of RGB images from one or more image capturing sensors of the ADV(1802). The system synchronizes the RGB images with the point clouds to obtain RGB point clouds(1803). The system extracts features from the RGB point clouds, the features including contextual and spatial information of the RGB point clouds(1804). The system registers the RGB point clouds based on the extracted features(1805) and generates a point cloud map based on the registration of the RGB point clouds(1806).

Classes IPC  ?

• G01C 21/00 - Navigation; Instruments de navigation non prévus dans les groupes
• G06T 3/40 - Changement d'échelle d'une image entière ou d'une partie d'image

Abrégé

A system identifies a road to be navigated by an ADV, the road being captured by one or more point clouds from one or more LIDAR sensors (2601). The system extracts road marking information of the identified road from the point clouds, the road marking information describing one or more road markings of the identified road (2602). The system partitions the road into one or more road partitions based on the road markings (2603). The system generates a point cloud map based on the road partitions, where the point cloud map is utilized to perceive a driving environment surrounding the ADV (2604).

Classes IPC  ?

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

Abrégé

According to one embodiment, a system receives, at a host system from a data processing (DP) accelerator, an accelerator identifier (ID) that uniquely identifies the DP accelerator), wherein the host system is coupled to the DP accelerator over a bus. The system transmits the accelerator ID to a predetermined trusted server over a network. The system receives a certificate from the predetermined trusted server over the network, the certificate certifying the DP accelerator. The system extracts a public root key (PK_RK) from the certificate for verification, the PK_RK corresponding to a private root key (SK_RK) associated with the DP accelerator. The system establishes a secure channel with the DP accelerator using the PK_RK based on the verification to exchange data securely between the host system and the DP accelerator.

Classes IPC  ?

• H04L 9/32 - Dispositions pour les communications secrètes ou protégées; Protocoles réseaux de sécurité comprenant des moyens pour vérifier l'identité ou l'autorisation d'un utilisateur du système

Abrégé

A data processing system performs a secure boot using a security module (e. g., a trusted platform module (TPM)) of a host system (301). The system verifies that an operating system (OS) and one or more drivers including an accelerator driver associated with a data processing (DP) accelerator is provided by a trusted source (302). The system launches the accelerator driver within the OS (303). The system generates a trusted execution environment (TEE) associated with one or more processors of the host system (304). The system launches an application and a runtime library within the TEE, where the application communicates with the DP accelerator via the runtime library and the accelerator driver (305).

Classes IPC  ?

• G06F 9/30 - Dispositions pour exécuter des instructions machines, p.ex. décodage d'instructions

Abrégé

A system receives, at a runtime library executed within a trusted execution environment (TEE) of a host system, a request from an application to invoke a predetermined function to perform a predefined operation. In response to the request, the system identifies a kernel object associated with the predetermined function. The system verifies an executable image of the kernel object using a public key corresponding to a private key that was used to sign the executable image of the kernel object. In response to successfully the system verifies the executable image of the kernel object, transmitting the verified executable image of the kernel object to a data processing (DP) accelerator over a bus to be executed by the DP accelerator to perform the predefined operation.

Classes IPC  ?

• G06F 21/51 - Contrôle des usagers, programmes ou dispositifs de préservation de l’intégrité des plates-formes, p.ex. des processeurs, des micrologiciels ou des systèmes d’exploitation au stade du chargement de l’application, p.ex. en acceptant, en rejetant, en démarrant ou en inhibant un logiciel exécutable en fonction de l’intégrité ou de la fiabilité de la source

Abrégé

According to one embodiment, a system receives, at a host system a public attestation key (PK_ATT) or a signed PK_ATT from a data processing (DP) accelerator over a bus. The system verifies the PK_ATT using a public root key (PK_RK) associated with the DP accelerator. In response to successfully verifying the PK_ATT, the system transmits a kernel identifier (ID) to the DP accelerator to request attesting a kernel object stored in the DP accelerator. In response to the system receives a kernel digest or a signed kernel digest corresponding to the kernel object form the DP accelerator, verifying the kernel digest using the PK_ATT. The system sends the verification results to the DP accelerator for the DP accelerator to access the kernel object based on the verification results.

Classes IPC  ?

• H04L 9/14 - Dispositions pour les communications secrètes ou protégées; Protocoles réseaux de sécurité utilisant plusieurs clés ou algorithmes

Abrégé

According to one embodiment, in response to receiving a temporary public key (PK_d) from a data processing (DP) accelerator, a system generates a first nonce (nc) at the host system, where the DP accelerator is coupled to the host system over a bus. The system transmits a request to create a session key from the host system to the DP accelerator, the request including a host public key (PK_O) and the first nonce. The system receives a second nonce (ns) from the DP accelerator, where the second nonce is encrypted using the host public key and a temporary private key (SK_d) corresponding to the temporary public key. The system generates a first session key based on the first nonce and the second nonce, which is utilized to encrypt or decrypt subsequent data exchanges between the host system and the DP accelerator.

Classes IPC  ?

• H04L 9/00 - Dispositions pour les communications secrètes ou protégées; Protocoles réseaux de sécurité

Abrégé

A DP accelerator includes one or more execution units (EUs) configured to perform data processing operations in response to an instruction received from a host system coupled over a bus. The DP accelerator includes a time unit (TU) coupled to the security unit to provide timestamp services. The DP accelerator includes a security unit (SU) configured to establish and maintain a secure channel with the host system to exchange commands and data associated with the data processing operations, where the security unit includes a secure storage area to store a private root key associated with the DP accelerator, where the private root key is utilized for authentication. The SU includes a random number generator to generate a random number, and a cryptographic engine to perform cryptographic operations on data exchanged with the host system over the bus using a session key derived based on the random number.

Classes IPC  ?

• H04L 9/32 - Dispositions pour les communications secrètes ou protégées; Protocoles réseaux de sécurité comprenant des moyens pour vérifier l'identité ou l'autorisation d'un utilisateur du système
• H04L 9/08 - Répartition de clés

Abrégé

A DP accelerator (405) includes one or more execution units (EUs) configured to perform data processing operations in response to an instruction received from a host system coupled over a bus. The DP accelerator (405) includes a security unit (1020) configured to establish and maintain a secure channel with the host system to exchange commands and data associated with the data processing operations. The DP accelerator (405) includes a time unit (2003) coupled to the security unit (1020) to provide timestamp services to the security unit (1020), where the time unit (2003) includes a clock generator to generate clock signals locally without having to derive the clock signals from an external source. The time unit includes a timestamp generator coupled to the clock generator to generate a timestamp based on the clock signals, and a power supply to provide power to the clock generator and the timestamp generator.

Classes IPC  ?

• G06F 21/72 - Protection de composants spécifiques internes ou périphériques, où la protection d'un composant mène à la protection de tout le calculateur pour assurer la sécurité du calcul ou du traitement de l’information dans les circuits de cryptographie

Abrégé

According to one embodiment, a system establishes a secure connection between a host system and a data processing (DP) accelerator over a bus, the secure connection including one or more data channels. The system transmits a first instruction from the host system to the DP accelerator over a command channel, the first instruction requesting the DP accelerator to perform a data preparation operation. The system receives a first request to read a first data from a first memory location of the host system from the DP accelerator over one data channel. In response to the request, the system transmits the first data to the DP accelerator over the data channel, where the first data is utilized for a computation or a configuration operation. The system transmits a second instruction from the host system to the DP accelerator over the command channel to perform the computation or the configuration operation.

Classes IPC  ?

• H04L 29/08 - Procédure de commande de la transmission, p.ex. procédure de commande du niveau de la liaison

Abrégé

According to one embodiment, a system performs a secure boot using a security module such as a trusted platform module (TPM) of a host system. The system establishes a trusted execution environment (TEE) associated with one or more processors of the host system. The system launches a memory manager within the TEE, where the memory manager is configured to manage memory resources of a data processing (DP) accelerator coupled to the host system over a bus, including maintaining memory usage information of global memory of the DP accelerator. In response to a request received from an application running within the TEE for accessing a memory location of the DP accelerator, the system allows or denies the request based on the memory usage information.

Classes IPC  ?

• G06F 21/74 - Protection de composants spécifiques internes ou périphériques, où la protection d'un composant mène à la protection de tout le calculateur pour assurer la sécurité du calcul ou du traitement de l’information opérant en mode dual ou compartimenté, c. à d. avec au moins un mode sécurisé

Abrégé

According to one embodiment, a system receives, at a host channel manager (HCM) of a host system, a request from an application to establish a secure channel with a data processing (DP) accelerator, where the DP accelerator is coupled to the host system over a bus. In response to the request, the system generates a first session key for the secure channel based on a first private key of a first key pair associated with the HCM and a second public key of a second key pair associated with the DP accelerator. In response to a first data associated with the application to be sent to the DP accelerator, the system encrypts the first data using the first session key. The system then transmits the encrypted first data to the DP accelerator via the secure channel over the bus.

Classes IPC  ?

• G06F 21/72 - Protection de composants spécifiques internes ou périphériques, où la protection d'un composant mène à la protection de tout le calculateur pour assurer la sécurité du calcul ou du traitement de l’information dans les circuits de cryptographie
• G06F 21/77 - Protection de composants spécifiques internes ou périphériques, où la protection d'un composant mène à la protection de tout le calculateur pour assurer la sécurité du calcul ou du traitement de l’information dans les cartes à puce intelligentes
• H04L 9/08 - Répartition de clés

Abrégé

An ADV perceives a driving environment surrounding the ADV based on sensor data obtained from a variety of sensors mounted on the ADV including, for example, perceiving and recognizing a corner the ADV may be about to turn. Based on the perception data of the driving environment, a set of features representing the characteristics of an entrance point of a corner that the ADV is about to turn. Based on the characteristics of the corner, an entrance point of the corner is determined. Based on the entrance point, a lookup operation is performed in a corner mapping table to locate a mapping entry matching the entrance point. A turning radius is then obtained from the mapping entry of the corner mapping table. The turning radius obtained from the corner mapping table is then utilized to plan a trajectory (e.g., steering angle) to drive the ADV to turn the corner.

Classes IPC  ?

• B60W 40/10 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés au mouvement du véhicule
• B62D 6/00 - Dispositions pour la commande automatique de la direction en fonction des conditions de conduite, qui sont détectées et pour lesquelles une réaction est appliquée, p.ex. circuits de commande
• G01C 21/32 - Structuration ou formatage de données cartographiques
• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions

Abrégé

A computer-implemented method for operating an autonomous driving vehicle is provided, the method comprising: perceiving a driving environment surrounding an autonomous driving vehicle (ADV), including identifying one or more objects based on sensor data obtained from a plurality of sensors of the ADV; for each of the identified objects, generating an arc curve connecting a current location of the ADV and the object, and calculating a curvature of the arc curve associated with the object; selecting one of the objects associated with an arc curve that satisfies a predetermined condition; and generating a reference line from the current location of the ADV to the selected object, wherein the reference line is utilized to generate a trajectory to drive the ADV. A non-transitory machine-readable medium and a data processing system is also provided.

Classes IPC  ?

• B60W 40/04 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés aux conditions ambiantes liés aux conditions de trafic

Abrégé

A three-point-turn is planned and executed in the operation of an autonomous driving vehicle (ADV). A candidate route from a start point and going through an end point is determined, the start point and the end point being in lanes associated with opposite travel directions. The candidate route is categorized into partially overlapping first, second, and third segments. A total cost associated with the candidate route is determined based at least in part on the first and second segments. Whether the total cost is below a threshold cost is determined. In response to a determination that the total cost is below the threshold cost, the three-point-turn is planned based on the candidate route. Further, driving signals are generated based at least in part on the planned three-point-turn to control operations of the ADV.

Classes IPC  ?

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

Abrégé

A system and a method for filtering obstacles to reduce the number of obstacles for an autonomous driving vehicle (ADV) to process in a given planning phase. The ADV can identify a first set of obstacles based on a set of criteria in a first lane where the ADV is travelling, filter out the remaining obstacles in the first lane, and expand each identified obstacle to a width of the first lane from the view of the ADV so that the ADV cannot nudge any of the first set of identified obstacles. When switching from the first lane to a second lane, the ADV can identify a second set of obstacles in the second lane using the same set of criteria, and expand each obstacle in the second set of obstacles to a width of the second lane while keep tracking the first set of identified obstacles. When the lane switching is completed, the ADV can stop tracking the first set of identified obstacles.

Classes IPC  ?

• B60W 30/14 - Régulateur d'allure

Abrégé

According to various embodiments, systems and methods are provided for use by an ADV to nudge incoming objects in a self-reverse lane. In an embodiment, an ADV determines that a self-reverse lane has a predetermined width before entering the self-reverse lane. After entering the self-reverse lane, the ADV can follow a first reference line at the center of the self-reverse lane. In response to detecting an incoming object, the ADV creates an alternative lane in the self-reverse lane by temporarily modifying a high definition map. The ADV subsequently follows a second reference line in the alternative lane to nudge the incoming object. In response to detecting that the incoming object has passed and the self-reverse lane is clear, the ADV can drive back to the center of the self-reverse lane, to continue to follow the first reference line in the self-reverse lane.

Classes IPC  ?

• B60W 30/08 - Anticipation ou prévention de collision probable ou imminente
• G08G 1/16 - Systèmes anticollision

Abrégé

Systems and methods for smoothing a reference line for an autonomous driving vehicle are described. In a method, a raw reference line can be generated from a high definition map and routing results, and can be truncated based on a predetermined formula. The truncated raw reference line can include a number of points, each point representing a position of an autonomous driving vehicle in a global coordinate system. The number of points can be converted to points in a local coordinate system, where a polynomial curve that best fits the points are generated. The polynomial curve can subsequently be used to generate a new vertical coordinate for a horizontal coordinate of each of the number of points. The new vertical coordinates and their corresponding horizontal coordinates can be converted back to the global coordinate system. The polynomial curve can be used to derive a heading, a kappa, and a dkappa for each point in the global coordinate system.

Classes IPC  ?

• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire

Abrégé

A computer-implemented method for operating an autonomous driving vehicle (ADV) is disclosed, the method comprising: determining a first torque value at a first time instant prior to executing a control command; determining a control command based on a speed of the ADV, a desired acceleration, and an associated entry in a calibration table; executing the control command; determining a second torque value at a second time instant subsequent to executing the control command; determining a torque error value as a difference between the first and second torque values; updating the associated entry in the calibration table based at least in part on the torque error value; and generating driving signals based at least in part on the updated calibration table to control operations of the ADV. A non-transitory machine-readable medium and a data processing system are also disclosed.

Classes IPC  ?

• B60K 31/00 - Accessoires agissant sur un seul sous-ensemble pour la commande automatique de la vitesse des véhicules, c. à d. empêchant la vitesse de dépasser une valeur déterminée de façon arbitraire ou maintenant une vitesse donnée choisie par le conducteur du

Abrégé

A path for parking is planned in operating an autonomous driving vehicle (ADV). The operations comprises: determining a plurality of sample points; determining a plurality of candidate paths connecting a start point and an end point for parking, each of the candidate paths passing through one of the sample points; determining a cost associated with each of the plurality of candidate paths; determining one or more candidate paths from the plurality of candidate paths that meet a boundary check requirement; selecting as the planned path a candidate path associated with a lowest cost that meets the boundary check requirement; determining a speed profile based on the planned path and an environment of the ADV; and generating driving signals based at least in part on the speed profile to control operations of the ADV to perform parking along the planned path.

Classes IPC  ?

• B60W 30/06 - Manœuvre automatique de stationnement
• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire

Abrégé

A pedestrian probability prediction system for autonomous vehicles (101) is disclosed. The system receives a captured image perceiving an environment of an ADV from an image capturing device of the ADV (801); the system identifies an obstacle in motion near the ADV based on the captured image (802); the system predicts a location for the moving obstacle at each of a number of time points (803); the system generates a probability ellipse based on the predicted location at the each time point (804), where the probability ellipse includes a probability indicator indicating different probabilities of the moving obstacle for different locations within the probability ellipse at the each time point.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• B60W 30/09 - Entreprenant une action automatiquement pour éviter la collision, p.ex. en freinant ou tournant

Abrégé

In response to a request for a three-point turn, a forward turning path from a current location and heading direction of the ADV is generated. In generating the forward turning path, a forward curvature is determined based on the maximum forward turning angle of the ADV by applying a full steering command. The forward turning path is determined based on the forward curvature from the current location of the ADV. A forward speed profile is calculated for the forward turning path based on perception information that perceives a driving environment surrounding the vehicle at the point in time. In addition, a backward turning path is generated from an end point of the forward turning path based on a maximum backward turning angle associated with the ADV. The three-point turn path is then generated based on the forward turning path and the backward turning path to drive the vehicle to make the three-point turn.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• G01C 21/00 - Navigation; Instruments de navigation non prévus dans les groupes

Abrégé

In response to a request for a three-point turn (701), a set of forward turning paths is generated based on a maximum forward turning angle associated with an ADV (702). A set of backward tuning paths is generated based on a maximum backward turning angle associated with the ADV (703). A set of three-point turn path candidates is generated based on the forward turning paths and the backward turning paths (704). For each of the three-point turn path candidates, a path cost is calculated using a predetermined cost function (705). One of the three-point turn path candidates with the lowest path cost is selected as the final three-point turn path to drive the ADV to make a three-point turn (706).

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire

Abrégé

In response to a request to make a three-point turn for an autonomous driving vehicle (ADV) (601), a forward turning path is generated using a first spiral function based on a maximum forward curvature change rate (602). A backward turning path is generated using a second spiral function based on a maximum backward curvature change rate (603). The forward and backward curvature change rates may be determined based on the maximum forward and backward turning angles associated with the ADV, which may be specified as a part of vehicle specification or vehicle design of the ADV. The backward turning path is initiated from an end point of the forward turning path (604). A three-point turn path is then generated based on the forward turning path and the backward turning path. The ADV is then driven according to the three-point turn path by issuing one or more proper control commands (605).

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• G01C 21/00 - Navigation; Instruments de navigation non prévus dans les groupes

Abrégé

A method, a non-transitory machine-readable medium and a data processing system for an autonomous driving vehicle. The system receives a captured image perceiving an environment of an ADV from an image capturing device of the ADV (601), where the captured image identifies an obstacle in motion near the ADV. The system generates a feasible area surrounding the moving obstacle based on a projection of the moving obstacle (602). If the ADV is within the feasible area, the system determines an upper bound velocity limit for the ADV (603). The system generates a trajectory having a trajectory velocity less than the upper bound velocity limit to control the ADV autonomously according to the trajectory such that if the ADV is within the feasible area the ADV is to decelerate (604).

Classes IPC  ?

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

Abrégé

According to one embodiment, a system receives a captured image perceiving an environment of an ADV from an image capturing device of the ADV capturing a plurality of obstacles near the ADV. The system generates a first tunnel based on a width of a road lane for the ADV, where the first tunnel represents a passable lane for the ADV to travel through. The system generates one or more additional tunnels based on locations of the obstacles, where the one or more additional tunnels modify a width of the passable lane according to a level of invasiveness of the obstacles. The system generates a trajectory of the ADV based on the first and the additional tunnels to control the ADV according to the trajectory to navigate around the obstacles without collision.

Classes IPC  ?

• B60W 30/14 - Régulateur d'allure
• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• B60W 30/18 - Propulsion du véhicule
• B60W 30/08 - Anticipation ou prévention de collision probable ou imminente
• B60W 40/06 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés aux conditions ambiantes liés à l'état de la route
• B60W 30/10 - Maintien de la trajectoire

Abrégé

A sensor unit (500) to be utilized in an autonomous driving vehicle (ADV) includes a sensor interface (504) that can be coupled to a number of sensors (510) mounted on a number of different locations of the ADV. The sensor unit (500) further includes a host interface (505) that can be coupled to a host system (110) such as a planning and control system utilized to autonomously drive the vehicle. The sensor unit (500) further includes a number of data transfer modules (502) corresponding to the sensors (510). Each of the data transfer modules (502) can be configured to operate in one of the operating modes, dependent uponthe type of the corresponding sensor. The operating modes include a low latency mode, a high bandwidth mode, and a memory mode.

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• H04B 3/54 - Systèmes de transmission par lignes de réseau de distribution d'énergie
• G06F 12/02 - Adressage ou affectation; Réadressage

Abrégé

A speed control command auto-calibration system for autonomous vehicles receives a first control command and a speed measurement of the ADV (901). The system determines an expected acceleration of the ADV based on the speed measurement and the first control command (902). The system receives an acceleration measurement of the ADV (903). The system determines a feedback error based on the acceleration measurement and the expected acceleration (904). The system updates a portion of the calibration table based on the determined feedback error (905). The system generates a second control command to control the ADV based on the calibration table having the updated portion to control the ADV autonomously according to the second control command (906).

Classes IPC  ?

• G05D 1/02 - Commande de la position ou du cap par référence à un système à deux dimensions
• G01C 21/34 - Recherche d'itinéraire; Guidage en matière d'itinéraire
• B60W 30/00 - Fonctions des systèmes d'aide à la conduite des véhicules routiers non liées à la commande d'un sous-ensemble particulier, p.ex. de systèmes comportant la commande conjuguée de plusieurs sous-ensembles du véhicule

Abrégé

Described herein are systems and methods for object detection to achieve hard real-time performance with low latency. Real-time object detection frameworks are disclosed. In one or more embodiments, a framework comprises a first CPU core, a second CPU core, and a plurality of shaves. In one or more embodiments, the first CPU core handles general CPU tasks, while the second CPU core handles the image frames from a camera sensor and computation task scheduling. In one or more embodiments, the scheduled computation tasks are implemented by the plurality of shaves using at least one object-detection model to detect an object in an image frame. In one or more embodiments, computation results from the object-detection model with a higher detection probability is used to form an output for object detection. In one or more embodiments, the object-detection models share some parameters for smaller size and higher implementing speed.

Classes IPC  ?

• G06T 1/20 - Architectures de processeurs; Configuration de processeurs p.ex. configuration en pipeline

Abrégé

Presented are systems and methods for improving speed and quality of real-time per-pixel depth estimation of scene layouts from a single image by using an end-to-end Convolutional Spatial Propagation Network (CSPN). An efficient linear propagation model performs propagation using a recurrent convolutional operation. The affinity among neighboring pixels may be learned through a deep convolutional neural network (CNN). The CSPN may be applied to two depth estimation tasks, given a single image: (1) to refine the depth output of existing methods, and (2) to convert sparse depth samples to a dense depth map, e.g., by embedding the depth samples within the propagation procedure. The conversion ensures that the sparse input depth values are preserved in the final depth map and runs in real-time and is, thus, well suited for robotics and autonomous driving applications, where sparse but accurate depth measurements, e.g., from LiDAR, can be fused with image data.

Classes IPC  ?

• G06T 7/50 - Récupération de la profondeur ou de la forme

Abrégé

Systems and methods are provided to improve the success rate of relocalization and eliminate the ambiguity of false relocalization by exploiting motions of the sensor system. During a relocalization process, a snapshot is taken using one or more visual sensors(110) and a single-shot relocalization in a visual map is implemented to establish candidate hypotheses. The visual sensors(110) move in the environment, with a movement trajectory tracked, to capture visual representations of the environment in one or more new poses. As the visual sensors(110) move, the relocalization system(100) tracks various estimated localization hypotheses and removes false ones until one winning hypothesis. Once the process is finished, the relocalization system(100) outputs a localization result with respect to the visual map.

Classes IPC  ?

• G01C 21/20 - Instruments pour effectuer des calculs de navigation

Abrégé

Various techniques for theft proofing autonomous driving vehicles (ADV) for transporting goods are described. Sensor data of a moving object representing a person within a predetermined proximity of an ADV for transporting goods (511) are captured for real-time analysis by a theft detection module, to determine a moving behavior of the moving object based on the sensor data in view of a set of known moving behaviors (513). The theft detection module further determines whether an intention of the person is likely to remove at least some of the goods from the ADV based on the moving behavior (515) using a process derived from historical image set, and sends an alarm to a predetermined destination in response to determining such an intention of the person. Other sensor data, for example, real time movements and weights of the ADV, can be used in conjunction with the process derived from historical image sets to determine the intention of the person.

Classes IPC  ?

• G08B 13/00 - Alarmes contre les cambrioleurs, les voleurs ou tous intrus
• B62D 63/02 - Véhicules à moteurs

Abrégé

A lateral drifting error is determined based on at least a current location of an ADV. The lateral drifting error is segmented into a first drifting error and a second drifting error using a predetermined segmentation algorithm. A planning module plans a path or trajectory for a current driving cycle to drive the ADV from the current location for a predetermined period of time. The planning module performs a first drifting error correction on the trajectory by modifying at least a starting point of the trajectory based on the first drifting error to generate a modified trajectory. A control module controls the ADV to drive according to the modified trajectory, including performing a second drifting error correction based on the second drifting error. As a result, the ADV can drive according to a path that is closer to an ideal situation.

Classes IPC  ?

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

Abrégé

Instead of using map data, a relative coordinate system is utilized to assist perception of the driving environment surrounding an ADV for some driving situations. One of such driving situations is driving on a highway. Typically, a highway has fewer intersections and exits. The relative coordinate system is utilized based on the relative lane configuration and relative obstacle information to control the ADV to simply follow the lane and avoid potential collision with any obstacles discovered within the road, without having to use map data. Once the relative lane configuration and obstacle information have been determined, regular path and speed planning and optimization can be performed to generate a trajectory to drive the ADV. Such a perception system is referred to as a relative perception system based on a relative coordinate system.

Classes IPC  ?

• B60Q 1/26 - Agencement des dispositifs de signalisation optique ou d'éclairage, leur montage, leur support ou les circuits à cet effet les dispositifs ayant principalement pour objet d'indiquer le contour du véhicule ou de certaines de ses parties, ou pour engendrer des signaux au bénéfice d'autres véhicules

Abrégé

A first localization system (401) performs a first localization using a first set of sensors (115A) to track locations of the ADV (300) along the path from a starting point to a destination point. A first localization curve (501) is generated as a result representing the locations of the ADV (300) along the path tracked by the first localization system (401). Currently, a second localization system (402) performs a second localization using a second set of sensors (115B) to track the locations of the ADV (300) along the path. A second localization curve (502) is generated as a result representing the locations of the ADV (300) along the path tracked by the second localization system (402). A system delay of the second localization system (402) is determined by comparing the second localization curve (502) against the first localization curve (501) as a localization reference. The system delay of the second localization system (402) can then be utilized to compensate path planning of the ADV (300) subsequently.

Classes IPC  ?

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