AI RESEARCH
Scene-Adaptive Continual Learning for CSI-based Human Activity Recognition with Mixture of Experts
arXiv CS.LG
•
ArXi:2605.06447v1 Announce Type: new Channel state information (CSI)-based human activity recognition (HAR) is vulnerable to performance degradation under domain shifts across varying physical environments. Continual learning (CL) offers a principled way to learn new domains sequentially while preserving past knowledge, but existing CL solutions for CSI-based HAR scale poorly with accumulating domains, rely on a large replay buffer, or incur linearly growing inference cost.