Attention-Based Sensor Fusion for Human Activity Recognition Using IMU Signals

• DOI: 10.48550/arxiv.2112.11224
• Publication Date: 2021-01-01
• Authors: Wenjin Tao; Haodong Chen; Md. Moniruzzaman; Ming C. Leu; Zhaozheng Yi; Ruwen Qin
• Keywords: Activity Recognition; Discriminative model; Sensor Fusion; Feature (linguistics)

Human Activity Recognition (HAR) using wearable devices such as smart watches embedded with Inertial Measurement Unit (IMU) sensors has various applications relevant to our daily life, workout tracking and health monitoring. In this paper, we propose a novel attention-based approach human activity recognition multiple IMU worn at different body locations. Firstly, sensor-wise feature extraction module is designed extract the most discriminative features from individual Convolutional Neural Networks (CNNs). Secondly, an fusion mechanism developed learn importance of locations generate attentive representation. Finally, inter-sensor applied correlations, which are connected classifier output predicted classes activities. The proposed evaluated five public datasets it outperforms state-of-the-art methods on wide variety categories.