TY  - RPRT
TI  - Multimodal Skeleton-Based Action Representation Learning via Decomposition and Composition
AU  - Hongsong Wang
AU  - Heng Fei
AU  - Bingxuan Dai
AU  - Jie Gui
PY  - 2025
DO  - 10.1007/s11633-025-1583-z
UR  - https://arxiv.org/abs/2512.21064
ID  - 2512.21064
ER  -