GaitCube: Deep Data Cube Learning for Human Recognition with Millimeter-Wave Radio

Abstract

Monitoring and identifying gait has recently emerged as a promising solution candidate for unobtrusive human recognition. In order to enable ubiquitous and reliable application, a gait recognition system must be robust to environment changes and easy to use without requiring too much user cooperation and recalibration, while maintaining high accuracy, which is often not satisfied in conventional approaches. In this paper, we present , a high-accuracy gait recognition system with minimal training requirement using a single commodity millimeter wave (mmWave) radio. To reduce the training overhead, we propose gait data cube, a novel 3D joint-feature representation of micro-Doppler and micro-Range signatures over time that can comprehensively embody the physical relevant features of one’s gait. With a pipeline of signal processing, can automatically detect and segment human walking and effectively extract the gait data cubes. We implement and evaluate through experiments conducted at 6 different locations in a typical indoor space with 10 subjects over a month, resulting in >50,000 gait instances. The results show that achieves an accuracy of 96.1% with a single gait cycle using one receive antenna, and the accuracy increases to 98.3% when combining all the receive antennas. Further, it achieves an average recognition accuracy of 79.1% for testing over different times and unseen locations by using only 2 minutes of training data collected in a single location, enabling a practical and ubiquitous gait-based identification.