Valeo Schalter Und Sensoren Gmbh

Jagdish Bhanushali

Cyril Rastoll

Adithya Ranga

Peter Groth

Sree Harsha Chowdary Gorantla

A method includes receiving image data that includes at least two images of an environment associated with a vehicle, identifying at least one person of interest in the image data, and generating, using a pose estimation model and the image data, a representation of the person of interest. The method also includes determining at least one characteristic associated with the at least two images of the image data and providing, to a machine learning model, at least the representation of the person of interest and the at least one characteristic associated with the at least two images of the image data. The method also includes receiving, from the machine learning model, a gesture prediction indicating a predicted gesture being made by the person of interest, and causing the vehicle to take at least one action based on the gesture prediction.

Systems and methods for person classification and gesture recognition

A method for generating an unobstructed bowl view of a vehicle that includes obtaining a plurality of images from a plurality of cameras disposed on the vehicle and determining a plurality of depth fields. The method further includes detecting a plurality of distorted objects in the plurality of images with a first machine-learned model that assigns a class distribution to the detected distorted object and estimating a distance of each distorted object from its associated camera using the plurality of depth fields. The method further includes assigning an object weight to each distorted object in the plurality of distorted objects and removing at least one distorted objects from the plurality of images. The method further includes replacing each of the at least one removed distorted objects with a representative background generated by a second machine-learned model and stitching and projecting the plurality of images to form the unobstructed bowl view.

Methods and systems for removing objects from view using machine learning

Methods, systems, computer-readable media, and apparatuses for blended control of remote operation of vehicles are presented. The sensors of automated vehicles can detect conditions for which there is no preprogrammed response. In these situations, automated vehicles can engage a remote operator mode. Remote operation of vehicles can be difficult. These techniques can use the automated driving system of the vehicle to monitor the driving commands from a remote operator and the sensor information of the vehicle. The commands from the remote operator can be analyzed by a supervision module in the automated vehicle. The supervision module can also receive vehicle sensor information. The supervision module can correct or adjust the remote operator commands as necessary to avoid collision, dampen remote operator induced oscillations, and provide for a smoother, safer, and more realistic driving experience.

Techniques for blended control for remote operations

Techniques are described for estimating intentions of pedestrians and other road users in vicinity of a vehicle. In certain embodiments, the techniques comprise obtaining, by a computer system of a vehicle equipped with one or more sensors, a sequence of video frames corresponding to a scene external to the vehicle, detecting one or more vulnerable road users (VRUs) in the sequence of video frames, wherein the detecting comprises estimating pose of each of the detected one or more VRUs. The techniques further include generating a segmentation map of the scene using one or more of the video frames; estimating one or more intention probabilities using estimated pose of the one or more VRUs and the segmentation map, each intention probability corresponding to one of the detected one or more VRUs, and adjusting one or more automated driving actions based on the estimated one or more intention probabilities.

Method and apparatus for predicting intent of vulnerable road users

Techniques are described for compressing image data for transmission to a computer system of a remote operator controlling a vehicle. A visual representation of a surrounding environment is generated from one or more images captured at the vehicle. One or more regions of interest in the visual representation are identified based at least on information received from the computer system of the remote operator, the information indicating a direction or area of focus of the remote operator. Regions of interest can be compressed to a lesser extent than regions located outside the regions of interest. The compressed visual representation is transmitted to the computer system for decompression and, ultimately, display on one or more display devices viewed by the remote operator.

Selective compression of image data during teleoperation of a vehicle

The present disclosure relates to a method, apparatus, and computer-readable storage medium for trajectory update using remote vehicle control. According to an embodiment, the above-described method, apparatus, and computer-readable storage medium of the present disclosure generally relate to identifying an obstacle blocking at least a portion of a road based on data received from one or more vehicle sensors, determining whether a path trajectory cannot be found to operate the vehicle with respect to the identified obstacle while the vehicle is operating in autonomous mode, sending, upon determination that the path trajectory cannot be found, a request to a remote controller to navigate the vehicle, operating the vehicle in remotely controlled mode, based upon instructions received from the remote controller, wherein a movement trajectory of the vehicle is recorded while being operated in the remotely controlled mode, and updating a navigational map based at least on the recorded movement trajectory.

Growing community of inventors

San Jose, CA, United States of America

Jagdish Bhanushali