Awesome Hand Pose Estimation is a curated reference collection for hand pose estimation research and resources.
The project addresses the need for researchers and practitioners to discover relevant work in hand pose estimation by organizing papers, datasets, and evaluation methods. It structures knowledge by categorizing papers across major computer vision conferences and journals spanning multiple years, separating work by input modality (depth-based, RGB-depth, and RGB), and providing links to datasets and evaluation benchmarks. The collection includes arXiv preprints, conference proceedings, journal articles, theses, and workshop materials alongside challenge competitions in the field.
Developers and researchers working on hand pose estimation, hand tracking, hand mesh recovery, or hand-object interaction should use this collection to survey the landscape of published approaches. It suits anyone building systems that require understanding hand articulation, keypoint detection, or 3D hand shape from visual input. The project is most valuable as a literature discovery tool rather than as implementation code, making it useful during the research and planning phases of projects involving hand analysis.
The project maintains an organized structure with dedicated sections for evaluation methodologies and separates datasets by their input characteristics. Contributions are explicitly welcomed, indicating an open approach to community participation in expanding the collection.
