This repository is a curated collection of research papers, open-source code, and learning resources related to Visual SLAM (Simultaneous Localization and Mapping) and Augmented Reality. It aims to track advancements in the field, particularly focusing on Visual Odometry (VO), object-level SLAM, and semantic data association, serving researchers and practitioners interested in these areas.

How It Works

The repository is structured into several key sections: "Open Source Code," "Excellent Authors and Labs," "SLAM Learning Resources," and "Recent Paper Updates." It meticulously lists and categorizes various SLAM algorithms and systems, providing links to their respective GitHub repositories, papers, and project pages. The "Recent Paper Updates" section is continuously updated with the latest research, offering a snapshot of current trends and developments.

Quick Start & Requirements

This repository does not provide a single install command as it is a collection of links to external projects. Each linked project will have its own installation and execution instructions, typically involving C++ (with dependencies like OpenCV, PCL, Eigen) or Python, and potentially requiring specific hardware like GPUs for deep learning-based methods.

Highlighted Details

• Comprehensive catalog of over 200 SLAM-related projects, including seminal works like ORB-SLAM2, DSO, LSD-SLAM, and SVO, as well as newer deep learning-based approaches.
• Detailed categorization of SLAM techniques, covering geometric, semantic, deep learning-based, sensor fusion (visual-inertial, LiDAR-visual-inertial), dynamic SLAM, and mapping.
• Extensive list of influential researchers and leading academic labs in the SLAM community, providing insights into key contributors and research directions.
• Curated learning resources, including tutorials, courses, and code explanations, to aid newcomers to the field.

Maintenance & Community

The repository was initiated in March 2019 and made public in March 2020. The last update mentioned was July 5, 2021. The author is graduating, indicating potential for infrequent updates. Contact is welcomed via email for academic exchange.

Licensing & Compatibility

The repository itself is a collection of links and does not have a specific license. However, the linked open-source projects will have their own licenses (e.g., GPL, MIT), which may have implications for commercial use or derivative works. Users must consult the license of each individual project.

Limitations & Caveats

The collection focuses on the author's specific interests (VO, object-level SLAM, semantic data association) and may not cover all aspects of visual SLAM. The last update was in July 2021, meaning more recent advancements might not be included. Some linked code repositories may be unmaintained or have complex build processes.