sjtuyinjie/awesome-LiDAR-Visual-SLAM

What it solves

Robots and autonomous systems need to know exactly where they are and what's around them. Two common sensors — LiDAR (laser-based) and cameras — each have strengths and weaknesses. LiDAR gives precise distance measurements but lacks visual detail; cameras provide rich color and texture but can be unreliable in poor lighting or motion. This project tackles that by collecting and organizing the most important research on combining both sensors into SLAM (Simultaneous Localization and Mapping) systems. Instead of hunting through scattered papers, this list gives a single, frequently updated overview of the best methods, complete with links to the original papers and open-source code.

How it works

The project is a hand-curated list, not a software tool. The author selects and categorizes top-tier publications on LiDAR-visual fusion SLAM, mostly from robotics and computer vision conferences and journals. Each entry includes the paper title, venue and year, and where available, direct links to the PDF, the official code repository, and sometimes datasets or demonstration videos. The author also highlights particularly notable works with a flame emoji and adds short parenthetical notes to explain what makes each method special. The list is ordered by year, from 2026 back to 2015, and the author invites other researchers to submit their own accepted works via pull requests. There is also a short introduction explaining why fusing LiDAR and camera data improves SLAM performance.

Who it's for

This is aimed at researchers, graduate students, and robotics engineers who work on SLAM, sensor fusion, or autonomous navigation. It's also useful for anyone who wants to learn about the state of the art in LiDAR-visual fusion without digging through dozens of separate sources. The clear links to code make it easy to try out a method or compare approaches.

Highlights

• Over 35 top-tier publications on LiDAR-visual fusion SLAM, spanning 2015 to 2026
• Each entry links to the official paper and, when available, to the source code and demo videos
• Highlights landmark systems like R3LIVE, FAST-LIVO2, and GS-LIVO with brief explanations of their innovations
• Covers both classic and recent approaches, including Gaussian-splatting-based SLAM and multi-camera systems
• Actively maintained, with a clear process for contributing new accepted works via GitHub pull requests
• Includes references to related awesome lists for wheel SLAM, Isaac Sim, and other robotics topics