A manifold learning approach for gesture recognition from micro-Doppler radar measurements

TL;DR

This paper explores a manifold learning method for gesture recognition from micro-Doppler radar data, demonstrating that localized kernels combined with PCA can achieve performance comparable to deep neural networks with faster training and lower memory use.

Contribution

It introduces a kernel-based manifold approximation approach that is domain-agnostic and effective for gesture recognition, with theoretical analysis and practical validation on radar and video data.

Findings

Localized kernel with PCA achieves near-DNN performance

Method offers faster training and lower memory requirements

Effective across different data domains

Abstract

A recent paper (Neural Networks, {\bf 132} (2020), 253-268) introduces a straightforward and simple kernel based approximation for manifold learning that does not require the knowledge of anything about the manifold, except for its dimension. In this paper, we examine how the pointwise error in approximation using least squares optimization based on similarly localized kernels depends upon the data characteristics and deteriorates as one goes away from the training data. The theory is presented with an abstract localized kernel, which can utilize any prior knowledge about the data being located on an unknown sub-manifold of a known manifold. We demonstrate the performance of our approach using a publicly available micro-Doppler data set, and investigate the use of different preprocessing measures, kernels, and manifold dimensions. Specifically, it is shown that the localized kernel introduced in the above mentioned paper when used with PCA components leads to a near-competitive performance to deep neural networks, and offers significant improvements in training speed and memory requirements. To demonstrate the fact that our methods are agnostic to the domain knowledge, we examine the classification problem in a simple video data set.