Summary: This repository is a curated bibliography of academic papers and datasets focused on egocentric (first-person) vision. It addresses the need for a centralized resource for researchers and engineers in this computer vision sub-field, covering areas like action recognition, object interaction, and pose estimation. Its benefit lies in providing a structured overview and quick access to key publications and data, accelerating discovery and research efforts.

How It Works: The repository organizes egocentric vision resources by problem statement (e.g., action recognition, hand-object interaction) and by major computer vision conferences (CVPR, ECCV, ICCV, WACV, BMVC). It lists academic papers with links to their sources and provides an extensive catalog of datasets, often with direct links to project pages or code repositories. This structure offers a comprehensive map of the egocentric vision research landscape.

Quick Start & Requirements: As a curated list, this repository requires no installation or software setup. Users should consult the individual papers and datasets linked within for their specific requirements, dependencies, and usage instructions.

Highlighted Details:

• Comprehensive coverage of egocentric vision sub-fields, including action recognition, object/hand recognition, pose estimation, gaze anticipation, and more.
• Papers are systematically organized by problem statement and major computer vision conferences, offering a structured view of research trends.
• An extensive catalog of egocentric datasets is provided, with many entries linking directly to code, project pages, and demos for immediate access to research artifacts.

Maintenance & Community: The repository is a "work in progress" and actively welcomes contributions, indicating a community-driven approach to maintaining and expanding its content.

Licensing & Compatibility: No license is specified for the curated list itself. Users must refer to the individual licenses of linked papers and datasets for terms of use and compatibility, especially for commercial applications.

Limitations & Caveats: This repository serves solely as a pointer to external research and data, offering no executable code or direct evaluation capabilities. Its completeness and up-to-dateness depend on ongoing community contributions.