This repository is a curated list of research papers on Multi-Modal Reinforcement Learning (MMRL), aimed at researchers and practitioners in AI and robotics. It provides a continuously updated collection of resources, focusing on agents that learn from diverse data sources like images and text, enabling more human-like intelligence and learning from easily accessible internet data.

How It Works

The repository organizes papers by major AI conferences (ICLR, NeurIPS, ICML, CVPR, CoRL) and ArXiv, facilitating targeted research. Each entry includes the paper title, link, authors, keywords, and the experimental environment used, offering a structured overview of the MMRL landscape.

Quick Start & Requirements

• No installation or execution required; this is a curated list of research papers.
• Links to papers, conference proceedings, and ArXiv are provided.

Highlighted Details

• Comprehensive coverage of MMRL research from 2017 to the latest conferences.
• Detailed categorization by conference and inclusion of keywords and experimental environments for each paper.
• Focus on agents learning from vision, language, or both, reflecting current AI trends.
• Includes papers that may be tangentially related but useful for MMRL research.

Maintenance & Community

• The repository is actively maintained and updated to track the frontier of MMRL.
• Contributions are welcomed, with instructions provided for interested parties.
• Links to community channels or social media are not specified.

Licensing & Compatibility

• Licensed under the Apache 2.0 license.
• Permissive license suitable for commercial use and integration into closed-source projects.

Limitations & Caveats

Some included papers may not be directly relevant to RL but are listed for their potential utility in MMRL research, requiring users to filter for direct applicability.