The problem of estimating instantaneous distributed target velocity using noisy measurements by multiple asynchronous automotive radar sensors is investigated. Two novel neural networks (NNs)-based approaches are proposed to address the problem. Both NNs use the point cloud with radar detections as an input. In the first approach, a hybrid NN is designed to take a set of points inside a cluster as its input, extract spatial-dynamic features to be used as weights for each input point, and apply them to obtain a weighted least square (WLS) solution for instantaneous velocity estimation. To this end, dedicated loss functions are proposed to allow the model to predict weights that can follow a velocity profile curve satisfying target constraints. Moreover, a small offset in the radial velocity value of each point is applied to adjust errors in the sensor measurements. In the second approach, a deep NN is proposed that takes a set of points inside a cluster as its input and directly outputs velocity estimates. Both approaches have been verified experimentally using the large open-source automotive RadarScenes dataset. The results show a significant improvement in terms of mean absolute error in velocity estimation over the state-of-the-art alternative techniques. Moreover, the estimated velocity is used as an additional measurement value inside a target tracker. Results show that this can increase the performance of the tracker, especially during challenging scenarios such as abrupt changes in the velocity of the target.

A data driven method is proposed to obtain free space segmentation using automotive radar point clouds. It aggregates automotive radar detection points from multiple timestamps, projects them into a Birds-Eye-View grid-based representation, and applies a semantic segmentation Neural Network (NN) to classify each grid for free space segmentation. A lidar based supervision is used to generate the ground truth for training. Moreover, debris objects are manually annotated to enable the NN model to learn to detect these uncommon objects. Experimental results on a proprietary 4D Imaging Radar dataset demonstrate that the proposed method gives improved free space segmentation as compared to other baseline methods.

A neural network (NN) based multi-frame classification approach is proposed to solve the problem of classification of tracked objects. Initially, a baseline tracker is implemented that uses the classification output of an object detection network for classification. Afterwards, two approaches for multi-frame classification are applied to perform classification of tracked objects. The first approach aggregates points from multiple frames and applies a single frame NN for classification, whereas the second approach uses bidirectional long short term memory (BiLSTM) layers to process points from multiple frames. Extensive experiments on the opensource 2D RadarScenes dataset showed a consistent increase in track performance when using either of the two techniques for multi-frame classification.

A novel temporal-spatial object classification neural network model is proposed to improve the classification capability of tracked objects. It takes queued points of tracked objects using multiple frames as input, utilizes spatial and temporal information from these points for sampling and grouping as well as extracts hierarchical temporal-spatial features for target classification. Experimental results on a proprietary 4D Imaging Radar dataset and open-source 2D RadarScenes dataset demonstrate that the proposed tracker-cued radar point-cloud target classification method allows the model to learn meaningful appearance and motion features from sparse radar points data, and achieves accurate classification output as compared to a baseline method, while being efficient to run on edge hardware with limited resources.

A single frame radar-based multi-object tracker that aims to improve data association for better tracking performance is proposed. Firstly, a baseline tracker based on track-by-detection paradigm was implemented for automotive radar. Secondly, investigation on the performance of the tracker when tracking individual classes separately versus all classes together was performed. Thirdly, appearance features were extracted from a neural network and added as an additional metric to the cost matrix for improved data association. Extensive experiments on the 2D RadarScenes dataset and a 3D proprietary Lunewave dataset (in partnership with NXP Semiconductors) showed a consistent improvement in the tracking performance using the approach proposed by adding features extracted from a neural network.