Toyota Motor Engineering & Manufacturing North America, Inc.

Toyota Jidosha Kabushiki Kaisha

Other

Toyota Motor Engineering & Manufacturing N.a. Inc.

Rui Guo

Hongsheng Lu

Kentaro Oguchi

Prashant Tiwari

Xuewei Qi

Ahmed Hamdi Sakr

Hao Yang

Sergei S Avedisov

Onur Altintas

Takayuki Shimizu

Kyungtae Han

Bin Cheng

Hansi Liu

Seyhan Ucar

Ziran Wang

Liu Liu

Shalini Keshavamurthy

Fei Chen

Haritha Muralidharan

Jianyu Su

Mahmoud Ashour

Kareem Metwaly

Takamasa Higuchi

Roger D Melen

BaekGyu Kim

John P Kenney

Gaurav Bansal

Baik Hoh

Yusuke Nakano

Chang-Heng Wang

Eric Martinson

Preeti Jayagopi Pillai

Qiang Liu

Mehmet Ali Guney

Peng Gao

Ippei Nishitani

Chang-heng Wang

Babajide Ayinde

Hao Sun

Wenjie Luo

Fangying Zhai

The disclosure includes embodiments for adaptive sensor data sharing by a connected vehicle. A method includes calculating, by a processor of an ego vehicle, a view angle overlap between a first sensor of the ego vehicle and a second sensor of a roadway device, wherein the view angle overlap is an amount of sensor view overlap shared by first sensor and the second sensor. The method includes determining an amount of ego sensor data to share with the roadway device based on the view angle overlap. The method includes building a sharing message that includes the amount of ego sensor data in a payload of the sharing message. The method includes transmitting, by a communication unit of the ego vehicle, the sharing message to the roadway device.

Adaptive sensor data sharing for a connected vehicle

System, methods, and other embodiments described herein relate to identifying connected vehicles through invoking maneuvers by vehicles that assist with distinguishing between safety messages having unassociated identifiers originating from the connected vehicles. In one embodiment, a method includes identifying surrounding vehicles including indeterminate vehicles from a driving environment using temporary identifiers received from the surrounding vehicles. The method also includes generating trajectories for the indeterminate vehicles. The method also includes distinguishing the indeterminate vehicles from communications using the temporary identifiers by observing execution of the trajectories by the indeterminate vehicles. The method also includes executing a planning task by a subject vehicle using safety information from the communications.

Systems and methods for identifying vehicles to communicate safety messages

Systems, methods, and other embodiments described herein relate to detecting obstacles from unknown objects during automated driving by a vehicle. In one embodiment, a method includes generating, from an image that includes an unknown object, a depth map from a depth estimation component and processed data from a semantic segmentation component in parallel by using a neural network model. The method also includes detecting that the unknown object is an obstacle when the unknown object satisfies criteria using an optical model according to the depth map and a segmentation map. The method also includes determining a height of the obstacle and a distance to the obstacle according to the optical model and the criteria. The method also includes adapting a vehicle plan of the automated driving according to the height.

Systems and methods for obstacle detection using a neural network model, depth maps, and segmentation maps

A vehicle includes one or more sensors configured to obtain raw data related to a scene, one or more processors, and machine readable instructions stored in one or more memory modules. The one machine readable instructions, when executed by the one or more processors, cause the vehicle to: process the raw data with a first neural network stored in the one or more memory modules to obtain a first prediction about the scene, transmit the raw data to a computing device external to the vehicle, receive a second prediction about the scene from the computing device in response to transmitting the raw data to the computing device, and determine an updated prediction about the scene based on a combination of the first prediction and the second prediction.

Methods and systems for enhanced scene perception using vehicle platoon

A method for measuring a distance is provided. The method includes receiving a fine time measurement (FTM) message, extracting channel state information (CSI) from the FTM message, determining with a channel condition classifier a probability that the FTM message is multi-path based on the corresponding CSI, and discarding the FTM message in response to the probability that the FTM message is multi-path is above a threshold level of probability.

Methods, systems, and non-transitory mediums for augmented fine time measurement

Systems and methods of cooperative autonomous driving which maximize objectives of a global traffic environment are provided. In particular, a cooperative controller may control multiple connected vehicles within a traffic environment. In certain embodiments, the cooperative controller may assign each connected vehicle a traffic role, based in part on their situation within the traffic environment. In some embodiments, this traffic role may also be based on a 'priority level' for the connected vehicle which corresponds to a desired travel time. Once traffic roles have been assigned, the cooperative controller may control each connected vehicle according to a driving policy associated with its assigned traffic role.

Cooperative artificial intelligent assisted driving

System, methods, and embodiments described herein relate to dynamically generating a wide field-of-view three-dimensional pseudo point cloud of an environment around a vehicle. A disclosed method may include capturing, via a camera, a first view in a first image, determining a first depth map based on the first image, obtaining, from an external system, a second image of a second view that overlaps the first view and a second depth map based on the second image, inputting the first image and the second image into a self-supervised homograph network that is trained to output a homographic transformation matrix between the first image and the second image, and generating a three-dimensional pseudo point cloud that combines the first depth map and the second depth map based on the homographic transformation matrix.

Cooperative pseudo point cloud for connected vehicles

A method includes calculating a first adaptive threshold based on uncertainty of a first parameter of a first collaborative perception message (CPM) from a first node and uncertainty of the first parameter of a second CPM from a second node, calculating a second adaptive threshold based on uncertainty of a second parameter of the first CPM and uncertainty of the second parameter of the second CPM, obtaining a first association matrix by filtering out one or more pairs whose score is greater than the first adaptive threshold, obtaining a second association matrix by filtering out one or more pairs whose score is greater than the second adaptive threshold, obtaining a fused association matrix based on the first association matrix and the second association matrix, and implementing a fusion algorithm on the fused associated matrix to obtain correspondence identification among objects.

Systems and methods for matching objects in collaborative perception messages using multiple adaptive thresholds

One embodiment of a system for analyzing the in-lane driving behavior of an external road agent generates a sequence of sparse 3D point clouds based on a sequence of depth maps corresponding to a sequence of images of a scene. The system performs flow clustering based on the sequence of depth maps and a sequence of flow maps to identify points across the sequence of sparse 3D point clouds that belong to a detected road agent. The system generates a dense 3D point cloud by combining at least some of those identified points. The system detects one or more lane markings and projects them into the dense 3D point cloud to generate an annotated 3D point cloud. The system analyzes the in-lane driving behavior of the detected road agent based, at least in part, on the annotated 3D point cloud.

Systems and methods for analyzing the in-lane driving behavior of a road agent external to a vehicle

Systems and methods are provided for utilizing sensor data from sensors of different modalities and from different vehicles to generate a combined image of an environment. Sensor data, such as a point cloud, generated by a LiDAR sensor on a first vehicle may be combined with sensor data, such as image data, generated by a camera on a second vehicle. The point cloud and image data may be combined to provide benefits over either data individually and processed to provide an improved image of the environment of the first and second vehicles. Either vehicle can perform this processing when receiving the sensor data from the other vehicle. An external system can also do the processing when receiving the sensor data from both vehicles. The improved image can then be used by one or both of the vehicles to improve, for example, automated travel through or obstacle identification in the environment.

Systems and methods of cooperative depth completion with sensor data sharing

System, methods, and other embodiments described herein relate to training a prediction system for improving depth perception in low-light. In one embodiment, a method includes computing, in a first training stage, losses associated with predicting a depth map for a synthetic image of a low-light scene, wherein the losses include a pose loss, a flow loss, and a supervised loss. The method also includes adjusting, according to the losses, a style model and a depth model. The method also includes training, in a second training stage, the depth model using a synthetic representation of a low-light image. The method also includes providing the depth model.

Systems and methods for training a prediction system

Described are systems and methods for substituting missing graph information caused during the transmission of graph information from one entity to another. In one example, a system includes a processor and a memory with machine-readable instructions that cause the processor to receive an incoming graph stream comprising a plurality of graphs having detected object information and at least one missing graph, transform the plurality of graphs into incoming embedded vectors having values that represent detected object information of the incoming graph stream, provide the embedded vector that corresponds to the at least one missing graph with a missing graph value, and substitute the missing graph value with a substitute value using a recurrent neural network that interpolates the substitute value using the incoming embedded vectors that are not equal to the missing graph value.

System and method for performing graph imputation

Connected vehicles and methods described herein provide for message sender identification. Connected vehicles and methods disclosed herein can include generating a message based hyper-graph based on the positions of connected elements of a road traffic network as communicated to the connected vehicle by the connected elements. Connected vehicles and methods disclosed herein can include generating, a perception based hyper-graph based on the perceived positions of at least a subset of the elements of the road traffic network as determined by a sensor of the vehicle. Connected vehicles and methods disclosed herein can include matching a node of the message-based hyper-graph to a corresponding node of the perception-based hyper-graph to determine the sender of a received message.

Vehicle communication sender identification via hyper-graph matching

A method for data sharing between a transmitter vehicle and a receiver vehicle includes obtaining sensor data, encoding the sensor data into structured data having a set of features and a set of channels, generating a channel attention map based on inter-channel relationships of the set of features, generating weights corresponding to channels of the channel attention map, selecting one or more channels among the set of channels based on the weights corresponding to the set of channels, and transmitting the selected channels to the receiver vehicle.

Deep cooperative feature sharing among connected vehicles

System, methods, and other embodiments described herein relate to training a model to stylize low-light images for improved perception. In one embodiment, a method includes encoding, by a style model, an input image to identify first content information. The method also includes decoding, by the style model, the first content information into an albedo component and a shading component. The method also includes generating, by the style model, a synthetic image using the albedo component and the shading component. The method also includes training the style model according to computed losses between the input image and the synthetic image.

Systems and methods for training a style model

Growing community of inventors

Mountain View, CA, United States of America

Rui Guo