Abstract: 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.

Abstract: 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.

Abstract: Systems and methods of the present disclosure provide an improved approach for open-set instance segmentation by identifying both known and unknown instances in an environment. For example, a method can include receiving sensor point cloud input data including a plurality of three-dimensional points. The method can include determining a feature embedding and at least one of an instance embedding, class embedding, and/or background embedding for each of the plurality of three-dimensional points. The method can include determining a first subset of points associated with one or more known instances within the environment based on the class embedding and the background embedding associated with each point in the plurality of points. The method can include determining a second subset of points associated with one or more unknown instances within the environment based on the first subset of points. The method can include segmenting the input data into known and unknown instances.

Abstract: Example aspects of the present disclosure describe a scene generator for simulating scenes in an environment. For example, snapshots of simulated traffic scenes can be generated by sampling a joint probability distribution trained on real-world traffic scenes. In some implementations, samples of the joint probability distribution can be obtained by sampling a plurality of factorized probability distributions for a plurality of objects for sequential insertion into the scene.

Abstract: 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.

Abstract: 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.

Abstract: 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.

Abstract: The present disclosure provides systems and methods for generating photorealistic image simulation data with geometry-aware composition for testing autonomous vehicles. In particular, aspects of the present disclosure can involve the intake of data on an environment and output of augmented data on the environment with the photorealistic addition of an object. As one example, data on the driving experiences of a self-driving vehicle can be augmented to add another vehicle into the collected environment data. The augmented data may then be used to test safety features of software for a self-driving vehicle.

Abstract: 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.

Abstract: 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.

Abstract: Systems and methods of the present disclosure provide an improved approach for open-set instance segmentation by identifying both known and unknown instances in an environment. For example, a method can include receiving sensor point cloud input data including a plurality of three-dimensional points. The method can include determining a feature embedding and at least one of an instance embedding, class embedding, and/or background embedding for each of the plurality of three-dimensional points. The method can include determining a first subset of points associated with one or more known instances within the environment based on the class embedding and the background embedding associated with each point in the plurality of points. The method can include determining a second subset of points associated with one or more unknown instances within the environment based on the first subset of points. The method can include segmenting the input data into known and unknown instances.

Abstract: 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.

Abstract: Systems and methods for determining a location based on image data are provided. A method can include receiving, by a computing system, a query image depicting a surrounding environment of a vehicle. The query image can be input into a machine-learned image embedding model and a machine-learned feature extraction model to obtain a query embedding and a query feature representation, respectively. The method can include identifying a subset of candidate embeddings that have embeddings similar to the query embedding. The method can include obtaining a respective feature representation for each image associated with the subset of candidate embeddings. The method can include determining a set of relative displacements between each image associated with the subset of candidate embeddings and the query image and determining a localized state of a vehicle based at least in part on the set of relative displacements.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices associated with object localization and generation of compressed feature representations are provided. For example, a computing system can access training data including a source feature representation and a target feature representation. An encoded target feature representation can be generated based on the target feature representation and a machine-learned encoding model. A binarized target feature representation can be generated based on the encoded target feature representation and lossless binarization operations. A reconstructed target feature representation can be generated based on the binarized target feature representation and a machine-learned decoding model. A matching score for the source feature representation and the reconstructed target feature representation can be determined. A loss associated with the matching score can be determined.

Abstract: 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.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices associated with object localization and generation of compressed feature representations are provided. For example, a computing system can access training data including a target feature representation and a source feature representation. An attention feature representation can be generated based on the target feature representation and a machine-learned attention model. An attended target feature representation can be generated based on masking the target feature representation with the attention feature representation. A matching score for the source feature representation and the target feature representation can be determined. A loss associated with the matching score and a ground-truth matching score for the source feature representation and the target feature representation can be determined. Furthermore, parameters of the machine-learned attention model can be adjusted based on the loss.

Abstract: Dense feature scale detection can be implemented using multiple convolutional neural networks trained on scale data to more accurately and efficiently match pixels between images. An input image can be used to generate multiple scaled images. The multiple scaled images are input into a feature net, which outputs feature data for the multiple scaled images. An attention net is used to generate an attention map from the input image. The attention map assigns emphasis as a soft distribution to different scales based on texture analysis. The feature data and the attention data can be combined through a multiplication process and then summed to generate dense features for comparison.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices associated with depth estimation are provided. For example, a feature representation associated with stereo images including a first and second plurality of points can be accessed. Sparse disparity estimates associated with disparities between the first and second plurality of points can be determined. The sparse disparity estimates can be based on machine-learned models that estimate disparities based on comparisons of the first plurality of points to the second plurality of points. Confidence ranges associated with the disparities between the first and second plurality of points can be determined based on the sparse disparity estimates and the machine-learned models. A disparity map for the stereo images can be generated based on using the confidence ranges and machine-learned models to prune the disparities outside the confidence ranges.

Abstract: 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.

Abstract: 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.