FlowGait: Enabling Robust Long-Term Gait Recognition Across Real-World Covariates with mmWave Radar

要旨

Gait recognition enables proactive and personalized smart home interactions, but its long-term reliability is challenged by the non-static nature of gait. Covariates like carrying items and clothing induce a persistent domain shift that degrades traditional, static models. To solve this, we introduce FlowGait, a mmWave-based framework designed for robust, long-term adaptation. It combines self-training with continual learning, allowing the model to daily align with a user's evolving gait by learning from readily available unlabeled data. It features a specialized transformer network for radar spectrogram analysis and a novel two-stage labeling algorithm that leverages the gait's hierarchical nature to assign pseudo-labels to the unlabeled data accurately. Evaluated on three challenging datasets from 47 volunteers (covering 12 gait-covariates, 11 routes, and two weeks), FlowGait achieves high accuracies of 94.8% (cross-covariate), 98.6% (cross-route), and 95.5% (cross-day). Notably, for the long-term dataset, it reduced performance decay from 13.6% to just 1.4%, demonstrating its real-world robustness.

著者
Dequan Wang
University of Science and Technology of China(USTC), HeFei, Anhui, China
Chenming He
University of Science and Technology of China, Hefei, China
Lingyu Wang
University of Science and Technology of China (USTC), Hefei, Anhui, China
Chengzhen Meng
University of Science and Technology of China, Hefei, Anhui, China
Xiaoran Fan
Independent Researcher, Sunnyvale, California, United States
Yanyong Zhang
University of Science and Technology of China (USTC), Hefei, Anhui, China

会議: CHI 2026

ACM CHI Conference on Human Factors in Computing Systems

セッション: Movement and Going Places

P1 - Room 134
7 件の発表
2026-04-17 20:15:00
2026-04-17 21:45:00