Including Semantic Information via Word Embeddings for Skeleton-based Action Recognition

Open in new window