Semantic-guided Cross-Modal Prompt Learning for Skeleton-based Zero-shot Action Recognition | IEEE Conference Publication | IEEE Xplore