Awesome Action Recognition

Aggregates hundreds of papers, code repositories, and datasets into a single curated list, saving researchers time from scouring disparate sources as evidenced by the extensive sections on video representation and datasets.

Organizes resources by subfields like temporal action detection and skeleton-based classification, making it easy to navigate specific research areas as shown in the table of contents and detailed subsections.

Actively updated via pull requests and email contributions per the README's guidelines, ensuring the list incorporates new state-of-the-art works and tools.

Includes direct links to code, project websites, and arXiv papers for each entry, such as for SlowFast Networks and ST-GCN, facilitating quick access to implementations.