Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing sensor data, e.g., laser sensor data, using neural networks. One of the methods includes obtaining a temporal sequence of multiple three-dimensional point clouds generated from sensor readings of an environment collected by one or more sensors within a given time period, each three-dimensional point cloud comprising a respective plurality of points in a first coordinate system; processing, using a feature extraction neural network, an input that comprises data derived from the temporal sequence of multiple three-dimensional point clouds to generate a feature embedding; receiving a query that specifies one time point within the given time period; and generating, from the feature embedding and conditioned on the query, one or more outputs that characterize one or more objects in the environment at the time point specified in the received query.

Abstract: An edge device can include a processing device, an image sensor, and a memory having instructions that are executable by the processing device for causing the processing device to perform operations. The processing device can receive, from the image sensor, an image of an environment. The processing device can determine a visibility measure corresponding to the environment by determining a dark channel of the image, determining, based on the dark channel of the image, a transmission map of the image, and determining, based on the transmission map, a visual contrast of the image. The processing device can generate information corresponding to the visibility measure.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output that characterizes a scene at a current time step. In one aspect, one of the systems include: a voxel neural network that generates a current early-stage feature representation of the current point cloud, a fusion subsystem that generates a current fused feature representation at the current time step; a backbone neural network that generates a current late-stage feature representation at the current time step, and an output neural network that generate an output that characterizes a scene at the current time step.

Abstract: A system receives a first request to execute a resource modification operation on any resources that meet a set of one or more resource selection criteria as of a time corresponding to the resource modification operation. The system stores the resource modification operation in association with the set of one or more resource selection criteria and the time. Responsive to detecting the time, the system executes a resource identification operation based on the set of one or more resource selection criteria to identify a resource, associated with a set of one or more attributes that satisfy the set of one or more resource selection criteria. Upon having identified the resource, the system initiates execution of the resource modification operation upon the resource to modify a configuration corresponding to the resource.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training an image representation neural network.

Abstract: Methods for quality control and optimizing the formation and characterization of micelles, vesicles or other aggregates are described herein. Pharmaceutically relevant peptides may be modified to form glycopeptide surfactants which form micelles or other aggregates with another surfactant. Glycopeptide and glycolipid surfactants can aggregate to form particles that enhance drug delivery. The glycopeptide surfactants may be drugs or prodrugs which are delivered via the micelles or other aggregated structures.

Abstract: A mobile roadway sensing unit can include an image sensor that can sense image data, a network sensor that can acquire wireless signal data, and a weather sensor that can sense weather condition data. The mobile roadway sensing unit can include a processor and a memory. The memory can include instructions executable by the processor for causing the processor to acquire the sensor data including at least one of the image data, the wireless signal data, or the weather condition data from the weather sensor, wherein the sensor data represents one or more current conditions proximate the mobile roadway sensing unit. The processor can generate processed data based on the sensor data The processor can generate, based on the processed data, a notification corresponding to the one or more current conditions. The processor can provide the notification to one or more external electronic devices.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training neural networks through contrastive learning. In particular, the contrastive learning is modified to use a relative margin to adjust a training pair's contribution to optimization.

Abstract: Methods, systems, and apparatus for processing point clouds using neural networks to perform a machine learning task. In one aspect, a system comprises one or more computers configured to obtain a set of point clouds captured by one or more sensors. Each point cloud includes a respective plurality of three-dimensional points. The one or more computers assign the three-dimensional points to respective voxels in a voxel grid, where the grid of voxels includes non-empty voxels to which one or more points are assigned and empty voxels to which no points are assigned. For each non-empty voxel, the one or more computers generate initial features based on the points that are assigned to the non-empty voxel. The one or more computers generate multi-scale features of the voxel grid, and the one or more computers generate an output for a point cloud processing task using the multi-scale features of the voxel grid.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for processing point cloud data to generate an output. In one aspect, a method includes obtaining a sequence of multiple point cloud frames, wherein the sequence of multiple point cloud frames comprise a target point cloud frame and a plurality of other point cloud frames; processing one or more other point cloud frames to generate one or more respective predicted locations at the target timestamp for each of one or more objects detected in the one or more other point cloud frames; generating, based on the respective predicted locations at the target timestamp for each of one or more objects, a synthetic point cloud frame that is associated with the target timestamp; and processing at least the synthetic point cloud frame to generate one or more outputs that characterize an environment at the target timestamp.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training neural networks through contrastive learning. In particular, the contrastive learning is modified to use a relative margin to adjust a training pair's contribution to optimization.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training neural networks through contrastive learning. In particular, the contrastive learning is modified to use a relative margin to adjust a training pair's contribution to optimization.

Abstract: The present application discloses a method and a system for analyzing and predicting a vehicle stay behavior based on multi-task learning, and the method includes the following steps: acquiring vehicle GPS and OBD data including a vehicle ID, a travel start time, a start longitude, a start latitude, an end time, an end longitude, and an end latitude after desensitization; preprocessing vehicle GPS and OBD data to obtain vehicle stay behavior data including stay location and stay duration; extract a spatial-temporal characteristic of the preprocessed vehicle stay behavior data by a deep recurrent neural network; inputting the spatial-temporal characteristic into a multi-task learning and predicting network, and obtaining the correlation between a stay location prediction task and the stay duration prediction task based on the historical stay behavior of the vehicle through the multi-task learning and predicting network to predict the stay location and stay duration.

Abstract: The present application discloses a method and a system for analyzing and predicting a vehicle stay behavior based on multi-task learning, and the method includes the following steps: acquiring vehicle GPS and OBD data including a vehicle ID, a travel start time, a start longitude, a start latitude, an end time, an end longitude, and an end latitude after desensitization; preprocessing vehicle GPS and OBD data to obtain vehicle stay behavior data including stay location and stay duration; extract a spatial-temporal characteristic of the preprocessed vehicle stay behavior data by a deep recurrent neural network; inputting the spatial-temporal characteristic into a multi-task learning and predicting network, and obtaining the correlation between a stay location prediction task and the stay duration prediction task based on the historical stay behavior of the vehicle through the multi-task learning and predicting network to predict the stay location and stay duration.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a machine learning model on training data. In one aspect, one of the methods include: obtaining a training data set comprising a plurality of training inputs; obtaining data defining an original search space of a plurality of candidate data augmentation policies; generating, from the original search space, a compact search space that has one or more global hyperparameters; and training the machine learning model on the training data using one or more final data augmentation policies generated from the compact search space.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output that characterizes a scene at a current time step. In one aspect, one of the systems include: a voxel neural network that generates a current early-stage feature representation of the current point cloud, a fusion subsystem that generates a current fused feature representation at the current time step; a backbone neural network that generates a current late-stage feature representation at the current time step, and an output neural network that generate an output that characterizes a scene at the current time step.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a machine learning model on training data. In one aspect, one of the methods include: obtaining a training data set comprising a plurality of training inputs; obtaining data defining an original search space of a plurality of candidate data augmentation policies; generating, from the original search space, a compact search space that has one or more global hyperparameters; and training the machine learning model on the training data using one or more final data augmentation policies generated from the compact search space.

Abstract: Embodiments are disclosed for computing accurate smooth occluding contours. In one embodiment, a method of computing accurate smooth occluding contours includes projecting a boundary polygon associated with a first region of a three-dimensional (3D) object to a two-dimensional (2D) image plane, the boundary polygon comprising a plurality of contour vertices and edges connecting the plurality of contour vertices, triangulating the first region in the 2D image plane to generate a 2D triangulation, and generating a 3D mesh for the first region by mapping the 2D triangulation to the 3D object.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing sensor data, e.g., laser sensor data, using neural networks. One of the methods includes obtaining a temporal sequence of multiple three-dimensional point clouds generated from sensor readings of an environment collected by one or more sensors within a given time period, each three-dimensional point cloud comprising a respective plurality of points in a first coordinate system; processing, using a feature extraction neural network, an input that comprises data derived from the temporal sequence of multiple three-dimensional point clouds to generate a feature embedding; receiving a query that specifies one time point within the given time period; and generating, from the feature embedding and conditioned on the query, one or more outputs that characterize one or more objects in the environment at the time point specified in the received query.

Abstract: An edge device can include a processing device, an image sensor, and a memory having instructions that are executable by the processing device for causing the processing device to perform operations. The processing device can receive, from the image sensor, an image of an environment. The processing device can determine a visibility measure corresponding to the environment by determining a dark channel of the image, determining, based on the dark channel of the image, a transmission map of the image, and determining, based on the transmission map, a visual contrast of the image. The processing device can generate information corresponding to the visibility measure.