The problem of instantaneous ego-motion estimation with mm-wave automotive radar is studied. DeepEgo, a deep learning-based method, is proposed for achieving robust and accurate ego-motion estimation. A hybrid approach that uses neural networks to extract complex features from input point clouds and applies weighted least squares (WLS) for motion estimation is utilized in DeepEgo. Additionally, a novel loss function, Doppler loss, is proposed to locate “inlier points” originating from detected stationary objects without human annotation. Finally, a challenging real-world automotive radar dataset is selected for extensive performance evaluation. Compared to other methods selected from the literature, significant improvements in estimation accuracy, long-term stability, and runtime performance of DeepEgo in comparison to other methods are demonstrated.@en

The problem of 2D instantaneous ego-motion estimation for vehicles equipped with automotive radars is studied. To leverage multi-dimensional radar point clouds and exploit point features automatically, without human engineering, a novel approach is proposed that transforms ego-motion estimation into a weighted least squares (wLSQ) problem using neural networks. Comparison with existing methods is done using a challenging real-world radar dataset. The comparison results show that the proposed method can achieve better performance in terms of estimation accuracy, long-term stability, and runtime performance compared to a representative approach selected from the recent literature.@en

The problem of estimating the 3D ego-motion velocity using multi-channel FMCW radar sensors has been studied. For the first time, the problem of ego-motion estimation is treated using radar raw signals. A robust algorithm using multi-channel FMCW radar sensors to instantly determine the complete 3D motion state of the ego-vehicle (i.e., translational speed and rotational speed) is proposed. The angle information of targets is extracted, and then their phase information from different times instances is used to determine vehicle ego-motion through an optimization process. Any pre-processing steps, such as clustering or clutter suppression, are not required. The performance of the algorithm is compared with the state-of-the-art algorithms based on real-world data, and superior performance has been demonstrated. The algorithm proposed can be easily integrated into radar signal processing pipelines for other tasks relevant to autonomous driving.@en

Linked to the increasing availability of datasets for radar-based human activity recognition (HAR), in this Student Highlights contribution, we report on a classification project that a group of 23 graduate students performed at TU Delft. The students were asked to work in groups of 2-3 members and to use the publicly available University of Glasgow dataset to develop the best classification pipeline as possible. This involved development and justification of both choices for the preprocessing techniques on the radar data (e.g., time-frequency distributions and cleaning of the signatures), and for the classification algorithms (e.g., the type of the algorithm, the hyperparameters' selection, the training-validation-testing split). While this student activity was performed at a small scale and with educational rather than research aims, we are happy to report it to the AESS readership, as we believe that such initiatives with open datasets sharing and classification algorithm benchmarking are beneficial for the wider radar research community. Furthermore, a list of publicly available datasets for radar-based HAR that can be used for similar initiatives is also reported in this article.

@en

Unconstrained human activities recognition with a radar network is considered. A hybrid classifier combining both convolutional neural networks (CNNs) and recurrent neural networks (RNNs) for spatial–temporal pattern extraction is proposed. The 2-D CNNs (2D-CNNs) are first applied to the radar data to perform spatial feature extraction on the input spectrograms. Subsequently, gated recurrent units with bidirectional implementations are used to capture the long- and short-term temporal dependencies in the feature maps generated by the 2D-CNNs. Three NN-based data fusion methods were explored and compared with utilize the rich information provided by the different radar nodes. The performance of the proposed classifier was validated rigorously using the K-fold cross-validation (CV) and leave-one-person-out (L1PO) methods. Unlike competitive research, the dataset with continuous human activities with seamless interactivity transitions that can occur at any time and unconstrained moving trajectories of the participants has been collected and used for evaluation purposes. Classification accuracy of about 90.8% is achieved for nine-class human activity recognition (HAR) by the proposed classifier with the halfway fusion method@en