This repository is a curated, comprehensive list of research papers focused on Video Object Segmentation (VOS). It serves as a valuable academic resource for researchers, engineers, and students seeking to explore the state-of-the-art in VOS, covering various sub-fields and organized by publication year.

How It Works

The repository functions as an academic bibliography, meticulously categorizing Video Object Segmentation research into Semi-Supervised, Unsupervised, and Referring VOS, alongside other related works. Each entry provides essential metadata such as paper titles, publication venues and years, and direct links to the original research papers (often via arXiv) and associated code repositories, facilitating easy access to the underlying methodologies.

Highlighted Details

• Extensive coverage of VOS literature from 2012 to the latest 2025 publications.
• Categorization into key VOS sub-fields: Semi-Supervised, Unsupervised, and Referring VOS.
• Inclusion of direct links to papers (arXiv, official pages) and, where available, code repositories for practical implementation.
• Chronological organization within categories aids in tracking the evolution of VOS techniques.

Maintenance & Community

The repository is maintained by suhwan-cho, with an open invitation for suggestions and requests, indicating a community-driven approach to expanding and refining the list. Specific community channels or detailed maintenance logs are not provided.

Licensing & Compatibility

No specific software license is declared for the curated list itself. The content comprises links to academic papers, each governed by its own publication and distribution terms. Compatibility for commercial use or closed-source linking would depend on the licenses of the individual research papers and their associated code.

Limitations & Caveats

This repository is a reference list and does not provide implementations, code, or direct access to the VOS models themselves. Users must independently locate, download, and implement the research papers listed. It also does not offer critical evaluations or summaries of the papers' methodologies or performance.