End-to-End Training of Neural Networks for Automotive Radar Interference Mitigation
This addresses interference mitigation for automotive radar systems, offering a novel training approach that improves detection accuracy.
The paper tackles the problem of mutual interference in automotive radar by training neural networks directly on object detection maps instead of regressing clean signals, resulting in a large margin increase in object detection performance.
In this paper we propose a new method for training neural networks (NNs) for frequency modulated continuous wave (FMCW) radar mutual interference mitigation. Instead of training NNs to regress from interfered to clean radar signals as in previous work, we train NNs directly on object detection maps. We do so by performing a continuous relaxation of the cell-averaging constant false alarm rate (CA-CFAR) peak detector, which is a well-established algorithm for object detection using radar. With this new training objective we are able to increase object detection performance by a large margin. Furthermore, we introduce separable convolution kernels to strongly reduce the number of parameters and computational complexity of convolutional NN architectures for radar applications. We validate our contributions with experiments on real-world measurement data and compare them against signal processing interference mitigation methods.