SAMoE-C achieves near state-of-the-art accuracy on cross-scene CSI HAR by selectively activating scene-specific experts via an attention router while using only a tiny replay buffer for training.
A model or 603 exemplars: Towards memory-efficient class-incremental learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Scene-Adaptive Continual Learning for CSI-based Human Activity Recognition with Mixture of Experts
SAMoE-C achieves near state-of-the-art accuracy on cross-scene CSI HAR by selectively activating scene-specific experts via an attention router while using only a tiny replay buffer for training.