Uatc, LLC

Aurora Operations, Inc.

Wei-Chiu Ma

Raquel Urtasun

Shenlong Wang

Ming Liang

Sivabalan Manivasagam

Bin Yang

Namdar Homayounfar

Yuwen Xiong

Wenyuan Zeng

Kelvin Ka Wing Wong

Shun Da Suo

Justin Liang

Rui Hu

Min Bai

Gellert Sandor Mattyus

Jerry Junkai Liu

Shrindihi Kowshika Lakshmikanth

Shivam Duggal

Ze Yang

Shrinidhi Kowshika Lakshmikanth

Generally, the disclosed systems and methods implement improved detection of objects in three-dimensional (3D) space. More particularly, an improved 3D object detection system can exploit continuous fusion of multiple sensors and/or integrated geographic prior map data to enhance effectiveness and robustness of object detection in applications such as autonomous driving. In some implementations, geographic prior data (e.g., geometric ground and/or semantic road features) can be exploited to enhance three-dimensional object detection for autonomous vehicle applications. In some implementations, object detection systems and methods can be improved based on dynamic utilization of multiple sensor modalities. More particularly, an improved 3D object detection system can exploit both LIDAR systems and cameras to perform very accurate localization of objects within three-dimensional space relative to an autonomous vehicle. For example, multi-sensor fusion can be implemented via continuous convolutions to fuse image data samples and LIDAR feature maps at different levels of resolution.

Three-dimensional object detection

The present disclosure provides systems and methods that combine physics-based systems with machine learning to generate synthetic LiDAR data that accurately mimics a real-world LiDAR sensor system. In particular, aspects of the present disclosure combine physics-based rendering with machine-learned models such as deep neural networks to simulate both the geometry and intensity of the LiDAR sensor. As one example, a physics-based ray casting approach can be used on a three-dimensional map of an environment to generate an initial three-dimensional point cloud that mimics LiDAR data. According to an aspect of the present disclosure, a machine-learned model can predict one or more dropout probabilities for one or more of the points in the initial three-dimensional point cloud, thereby generating an adjusted three-dimensional point cloud which more realistically simulates real-world LiDAR data.

Systems and methods for generating synthetic sensor data via machine learning

Systems, methods, tangible non-transitory computer-readable media, and devices associated with motion flow estimation are provided. For example, scene data including representations of an environment over a first set of time intervals can be accessed. Extracted visual cues can be generated based on the representations and machine-learned feature extraction models. At least one of the machine-learned feature extraction models can be configured to generate a portion of the extracted visual cues based on a first set of the representations of the environment from a first perspective and a second set of the representations of the environment from a second perspective. The extracted visual cues can be encoded using energy functions. Three-dimensional motion estimates of object instances at time intervals subsequent to the first set of time intervals can be determined based on the energy functions and machine-learned inference models.

Deep structured scene flow for autonomous devices

Systems and methods for generating simulation data based on real-world dynamic objects are provided. A method includes obtaining two- and three-dimensional data descriptive of a dynamic object in the real world. The two- and three-dimensional information can be provided as an input to a machine-learned model to receive object model parameters descriptive of a pose and shape modification with respect to a three-dimensional template object model. The parameters can represent a three-dimensional dynamic object model indicative of an object pose and an object shape for the dynamic object. The method can be repeated on sequential two- and three-dimensional information to generate a sequence of object model parameters over time. Portions of a sequence of parameters can be stored as simulation data descriptive of a simulated trajectory of a unique dynamic object. The parameters can be evaluated by an objective function to refine the parameters and train the machine-learned model.

Systems and methods for simulating dynamic objects based on real world data

The present disclosure provides systems and methods that combine physics-based systems with machine learning to generate synthetic LiDAR data that accurately mimics a real-world LiDAR sensor system. In particular, aspects of the present disclosure combine physics-based rendering with machine-learned models such as deep neural networks to simulate both the geometry and intensity of the LiDAR sensor. As one example, a physics-based ray casting approach can be used on a three-dimensional map of an environment to generate an initial three-dimensional point cloud that mimics LiDAR data. According to an aspect of the present disclosure, a machine-learned geometry model can predict one or more adjusted depths for one or more of the points in the initial three-dimensional point cloud, thereby generating an adjusted three-dimensional point cloud which more realistically simulates real-world LiDAR data.

Systems and methods for generating synthetic light detection and ranging data via machine learning

Systems and methods are provided for machine-learned models including convolutional neural networks that generate predictions using continuous convolution techniques. For example, the systems and methods of the present disclosure can be included in or otherwise leveraged by an autonomous vehicle. In one example, a computing system can perform, with a machine-learned convolutional neural network, one or more convolutions over input data using a continuous filter relative to a support domain associated with the input data, and receive a prediction from the machine-learned convolutional neural network. A machine-learned convolutional neural network in some examples includes at least one continuous convolution layer configured to perform convolutions over input data with a parametric continuous kernel.

Continuous convolution and fusion in neural networks

Disclosed herein are methods and systems for performing instance segmentation that can provide improved estimation of object boundaries. Implementations can include a machine-learned segmentation model trained to estimate an initial object boundary based on a truncated signed distance function (TSDF) generated by the model. The model can also generate outputs for optimizing the TSDF over a series of iterations to produce a final TSDF that can be used to determine the segmentation mask.

High quality instance segmentation

Systems and methods for generating object segmentations across videos are provided. An example system can enable an annotator to identify objects within a first image frame of a video sequence by clicking anywhere within the object. The system processes the first image frame and a second, subsequent, image frame to assign each pixel of the second image frame to one of the objects identified in the first image frame or the background. The system refines the resulting object masks for the second image frame using a recurrent attention module based on contextual features extracted from the second image frame. The system receives additional user input for the second image frame and uses the input, in combination with the object masks for the second image frame, to determine object masks for a third, subsequent, image frame in the video sequence. The process is repeated for each image in the video sequence.

Growing community of inventors

Toronto, Canada

Wei-Chiu Ma