Pure-Rust visual SLAM library visloc-rs benchmarks 3.46x faster than COLMAP on ETH3D
rsasaki0109 · x · 2026-09-18
visloc-rs is a pure-Rust implementation of visual and visual-inertial SLAM, SfM, and GPS-denied localization, aimed at robots and UAVs.
Author-reported benchmarks on ETH3D Electro (same input, 8-core CPU, end-to-end vs COLMAP 3.9.1):
- 27:29 vs 1:35:05 runtime — 3.46x faster
- Registered cameras: 1200/1200, parity with COLMAP
- Camera-centre RMSE: 3.50cm vs 4.68cm (25.2% lower)
- Courtyard dataset RMSE: 0.5379cm vs 1.6166cm (66.7% lower)
The repo (55 stars) includes benchmarks, examples, an optional CUDA backend for DPVO, and dual Apache/MIT licensing.
Related event: visloc-rs: Pure Rust Visual SLAM Goes Open Source(2 posts)→
More from Embodied
- Two Agentic Robots Clean a Room Together, Coordinating by Voice — chris_j_paxton · 2026-09-18
- Custom BB-SLAM hits 0.02m accuracy with pure visual odometry, no LiDAR or IMU — broodsugar · 2026-09-18
- DARPA surgical AI competition kicks off, veterans liken it to the 2004 self-driving challenge — Laparoscopes · 2026-09-18
- visloc-rs: Pure Rust library for visual & visual-inertial SLAM, SfM and localization — rsasaki0109 · 2026-09-18
- Scoble Teases Camera-Free, Display-Free All-Day AI Glasses Launching Next Week — Scobleizer · 2026-09-18
- VA-Bench: Best MLLM Scores Only 53.9% Task Success on Embodied Spatial Intelligence — dalian-university-of-technology · 2026-09-18