Attention‐enhanced Alexnet for improved radar micro‐Doppler signature classification. Issue 4 (30th December 2022)

• Record Type: Journal Article
• Authors: Vishwakarma, Shelly; Li, Wenda; Tang, Chong; Woodbridge, Karl; Adve, Ravi Raj; Chetty, Kevin
• Is Part Of: IET radar, sonar & navigation. Volume 17:Issue 4(2023)
• Page Start: 652
• Page End: 664
• Publication Date: 2022-12-30
• Subjects: image classification -- micro Doppler -- radar target recognition; Signal processing -- Periodicals; Radar -- Periodicals; Sonar -- Periodicals; Electronics in navigation -- Periodicals; Navigation -- Periodicals; 621.3848
• DOI: 10.1049/rsn2.12369
• Languages: English
• ISSNs: 1751-8784

Abstract

Abstract: This work introduces an attention mechanism that can be integrated into any standard convolution neural network to improve model sensitivity and prediction accuracy with minimal computational overhead. The attention mechanism is introduced in a lightweight network – Alexnet and its classification performance for human micro‐Doppler signatures is evaluated. The Alexnet model trained with an attention module can implicitly highlight the salient regions in the radar signatures while suppressing the irrelevant background regions and consistently improving network predictions. Network visualizations are provided through class activation mapping, providing better insights into how the predictions are made. The visualizations demonstrate how the attention mechanism focusses on the region of interest in the radar signatures. Abstract : This work improves model sensitivity and prediction accuracy with minimal computational overhead by introducing an attention mechanism into a standard convolution neural network. Moreover, it supports explainable deep learning through class activation map visualizations.