19 December 2023 in IEEE Transactions on Aerospace and Electronic Systems
Over the past few years, radar-based human activity recognition (HAR) has emerged as a prominent research area. The typical approach involves converting radar data into image data, such as micro-Doppler images, and then feeding them into networks for activity classification. However, relying solely on a single type of feature may restrict the ability of the networks to recognize activities accurately. To address this limitation and make the most of spatiotemporal features in the data, a dual-stream spatial and temporal feature fusion (DSTFF) network that leverages attention mechanism is proposed, which comprises a temporal feature extraction patial feature extraction stream (SFES) network. Additionally, a coordinates-based spatial attention mechanism (CSAM) is introduced to enhance the accuracy and efficiency of extracting deep spatial features. It focuses on the key spatial information from both horizontal and vertical directions of the feature map, which means range and velocity information in reality, respectively. Furthermore, it effectively connects the information across each channel. Meanwhile, a radar HAR (RadHAR) dataset based on two-dimensional (2D) and three-dimensional (3D) data is created by using a millimeter-wave radar. The evaluation experiments of the CSAM and the DSTFF network are carried out on the public dataset collected by the University of Glasgow (UOG) and the RadHAR dataset, respectively. The experimental results show that the CSAM exhibits strong generalization ability, and the DSTFF network presents effectiveness and superiority with the accuracy of 97.10%, which is higher than those of the classical and state-of-the-art networks.
11 November 2024 in IEEE Internet of Things Journal
Due to the flourishing of Internet of Things (IoT) technology, radar-based human activity recognition (HAR) technology has made significant progress and has become an indispensable research area. Many radar systems use multiple feature maps and handle them directly in image format. However, the generation of multiple forms of feature maps requires heavy computational resources, which makes it impractical for real-world applications. Additionally, many networks fail to fully extract temporal information from the time-Doppler (TD) map during the feature extraction. Therefore, a time-weighted network based on strip pooling (TWN-SP) using the TD map is proposed in this article. The TWN-SP consists of two time-weighted modules based on fire (TWMs-F) and a feature fusion module based on temporal and channel attention (FFM-TCA). Due to the application of depthwise separable convolution (DSC) and placing the feature fusion step at the forefront, the proposed network has fewer parameters. Moreover, a radar dataset named RadSet is constructed, containing TD maps of six daily activities. To validate the performance of the TWN-SP, a ten-fold cross-validation (CV) on the public dataset named Radar848 and a leave-one-subject-out (LOSO) CV on the RadSet dataset are carried out, respectively. To further validate the generalization performance of TWN-SP, a comparative analysis was conducted on another public dataset, Ci4R. The TWN-SP achieves an accuracy of 99.28% on the RadSet dataset. Experimental results indicate that the TWN-SP surpasses the current leading networks in both performance and complexity.