Safety
Frequency-Enhanced Diffusion Models: Curriculum-Guided Semantic Alignment for Zero-Shot Skeleton Action Recognition
arXiv:2604.09063v1 Announce Type: cross Abstract: Human action recognition is pivotal in computer vision, with applications ranging from surveillance to human-robot interaction. Despite the effectiven
arXiv:2604.09063v1 Announce Type: cross Abstract: Human action recognition is pivotal in computer vision, with applications ranging from surveillance to human-robot interaction. Despite the effectiveness of supervised skeleton-based methods, their reliance on exhaustive annotation limits generalization to novel actions. Zero-Shot Skeleton Action Recognition (ZSAR) emerges as a promising paradigm, yet it faces challenges due to the spectral bias of diffusion models, which oversmooth high-frequency dynamics. Here, we propose Frequency-Aware Diffusion for Skeleton-Text Matching (FDSM), integrating a Semantic-Guided Spectral Residual Module, a Timestep-Adaptive Spectral Loss, and Curriculum-based Semantic Abstraction to address these challenges. Our approach effectively recovers fine-grained motion details, achieving state-of-the-art performance on NTU RGB+D, PKU-MMD, and Kinetics-skeleton datasets. Code has been made available at https://github.com/yuzhi535/FDSM. Project homepage: https://yuzhi535.github.io/FDSM.github.io/
Related
- Commanding Humanoid by Free-form Language: A Large Language Action Model with Unified Motion Vocabulary
- Re-Mask and Redirect: Exploiting Denoising Irreversibility in Diffusion Language Models
- FP4 Explore, BF16 Train: Diffusion Reinforcement Learning via Efficient Rollout Scaling
- Visually-Guided Policy Optimization for Multimodal Reasoning
- Traj2Action: A Co-Denoising Framework for Trajectory-Guided Human-to-Robot Skill Transfer
- Commanding Humanoid by Free-form Language: A Large Language Action Model with Unified Motion Vocabulary
Source: arXiv cs.AI | 2026-04-13