Abstract: Upon identifying image data associated with an object, radar data associated with the object is identified. A semantic point cloud image is generated based on the image data and the radar data. Transformed semantic point cloud image from a perspective of the object is determined with a variational auto-encoder neural network trained to accept the semantic point cloud image of the object and to generate the transformed semantic point cloud image from the perspective of the object. Physical characteristics of the object are determined based on the transformed semantic point cloud image.

Abstract: Upon identifying image data associated with an object, radar data associated with the object is identified. A semantic point cloud image is generated based on the image data and the radar data. Transformed semantic point cloud image from a perspective of the object is determined with a variational auto-encoder neural network trained to accept the semantic point cloud image of the object and to generate the transformed semantic point cloud image from the perspective of the object. Physical characteristics of the object are determined based on the transformed semantic point cloud image.

Abstract: A system and method for localization includes a processing system with at least one processing device. The processing system is configured to obtain sensor data from a sensor system. The sensor system includes at least one radar sensor. The processing system is configured to obtain map data. The processing system is configured to determine if there is a predetermined number of detected features. The detected features are associated with the sensor data of a current sensing region of the sensor system. The processing system is configured to generate localization data based on the detected features of the current sensing region upon determining that the predetermined number of detected features is satisfied. The processing system is configured to obtain tracked feature data upon determining that the predetermined number of detected features is not satisfied and generate localization data based on the tracked feature data and the detected features of the current sensing region.

Abstract: A system and method for detecting a level of misalignment of a vehicle sensor, such as a radar sensor, compared to a reference sensor also associated with the vehicle. The reference sensor may comprise a different type of sensor, such as an optical sensor, motion sensor, or position sensor. One or more reference sensors may be utilized in detecting a level of misalignment.

Abstract: A method includes obtaining an occupancy grid. The method includes generating a reference, such as a ray, on the occupancy grid. The ray originates from a location of a vehicle and extends in a direction that corresponds to a driving direction of the vehicle. The method includes identifying markers along the ray. The markers include transition markers, which indicate a first type of transition or a second type of transition. The first type of transition is from an unoccupied zone to an occupied zone. The second type of transition is from the occupied zone to the unoccupied zone. The method includes generating a representation of drivable space in real-time based on computations involving sections that are associated with the first type of transition and the second type of transition. The method includes providing guidance to the vehicle based on the representation of drivable space.

Abstract: A method includes obtaining an occupancy grid. The method includes generating a reference, such as a ray, on the occupancy grid. The ray originates from a location of a vehicle and extends in a direction that corresponds to a driving direction of the vehicle. The method includes identifying markers along the ray. The markers include transition markers, which indicate a first type of transition or a second type of transition. The first type of transition is from an unoccupied zone to an occupied zone. The second type of transition is from the occupied zone to the unoccupied zone. The method includes generating a representation of drivable space in real-time based on computations involving sections that are associated with the first type of transition and the second type of transition. The method includes providing guidance to the vehicle based on the representation of drivable space.

Abstract: A system and method for detecting a level of misalignment of a vehicle sensor, such as a radar sensor, compared to a reference sensor also associated with the vehicle. The reference sensor may comprise a different type of sensor, such as an optical sensor, motion sensor, or position sensor. One or more reference sensors may be utilized in detecting a level of misalignment.

Abstract: A system and method for localization includes a processing system with at least one processing device. The processing system is configured to obtain sensor data from a sensor system. The sensor system includes at least one radar sensor. The processing system is configured to obtain map data. The processing system is configured to determine if there is a predetermined number of detected features. The detected features are associated with the sensor data of a current sensing region of the sensor system. The processing system is configured to generate localization data based on the detected features of the current sensing region upon determining that the predetermined number of detected features is satisfied. The processing system is configured to obtain tracked feature data upon determining that the predetermined number of detected features is not satisfied and generate localization data based on the tracked feature data and the detected features of the current sensing region.

Abstract: A system and method for determining fractional fat content of tissue comprises registering thermoacoustic image coordinates to an acquired ultrasound image, the acquired ultrasound image at least comprising target tissue within a region of interest; defining a thermoacoustic voxel grid coincident with the region of interest; obtaining thermoacoustic image measurement values from tissue within the region of interest corresponding to the voxels within the defined thermoecoustic voxel grid to yield a thermoacoustic measurement matrix; normalizing the thermoacoustic image measurement values within the thermoacoustic measurement matrix; calculating a fractional fat content map for the target tissue within the region of interest based on the normalized thermoacoustic image measurement values within the thermoacoustic measurement matrix and a reference thermoacoustic measurement value; and correcting the fractional fat content map based on tissue speed-of-sound data to yield a final fractional fat content map for th

Abstract: A system and method for determining fractional fat content of tissue comprises registering thermoacoustic image coordinates to an acquired ultrasound image, the acquired ultrasound image at least comprising target tissue within a region of interest; defining a thermoacoustic voxel grid coincident with the region of interest; obtaining thermoacoustic image measurement values from tissue within the region of interest corresponding to the voxels within the defined thermoecoustic voxel grid to yield a thermoacoustic measurement matrix; normalizing the thermoacoustic image measurement values within the thermoacoustic measurement matrix; calculating a fractional fat content map for the target tissue within the region of interest based on the normalized thermoacoustic image measurement values within the thermoacoustic measurement matrix and a reference thermoacoustic measurement value; and correcting the fractional fat content map based on tissue speed-of-sound data to yield a final fractional fat content map for th

Abstract: A method for correcting fat-induced aberrations in ultrasound imaging comprises segmenting a thermoacoustic absorption image of a region of interest into at least one fat region and at least one non-fat region, creating a speed of sound map by assigning a speed of sound to each region based on tissue type of the region, correcting aberrations in the segmented thermoacoustic absorption image using the assigned speeds of sound thereby generating a corrected thermoacoustic image, and correcting an ultrasound image of the region of interest using the corrected thermoacoustic image and the speed of sound map.

Abstract: A method for correcting fat-induced aberrations in ultrasound imaging comprises segmenting a thermoacoustic absorption image of a region of interest into at least one fat region and at least one non-fat region, creating a speed of sound map by assigning a speed of sound to each region based on tissue type of the region, correcting aberrations in the segmented thermoacoustic absorption image using the assigned speeds of sound thereby generating a corrected thermoacoustic image, and correcting an ultrasound image of the region of interest using the corrected thermoacoustic image and the speed of sound map.