[ICCV2025] "Bridging the Skeleton-Text Modality Gap: Diffusion-Powered Modality Alignment for Zero-shot Skeleton-based Action Recognition"

Jeonghyeok Do, Munchurl Kim

[Project Page]