microSLAM from Zurich team tops LaMaria benchmark using a single RGB stream
k7agar · x · 2026-09-11
A computer vision team in Zurich, with ETH Zurich's Computer Vision and Geometry Lab, announced microSLAM, a monocular SLAM system ranking first on the LaMaria benchmark while running on a single RGB stream — rivaling systems with many more cameras and IMUs. They argue robot learning is bottlenecked on data and head-mounted camera footage of real work is the largest untapped source, but only if geometry can be recovered from one moving camera in a dynamic world.
Related event: microSLAM tops LaMaria benchmark using only monocular RGB(2 posts)→
More from Embodied
- Unitree CEO Wang Xingxing approves every expense over 100 yuan: inside the robot maker's cost obsession — Afinetheorem · 2026-09-11
- Unitree's Playbook: Massive Scale, Low Costs, and R&D Only Where It Cuts Prices — Afinetheorem · 2026-09-11
- YC Demo Day: Live Drone Tracking Used by Ukraine's Military and Self-Improving Voice Agents — ycombinator · 2026-09-11
- OSU Professor Hai-Jun Su Wins Three ASME Awards for AI-Assisted Machine and Robot Design — AndrewDai · 2026-09-11
- Google AI teases robotics as its next frontier after finance — GoogleAI · 2026-09-11
- Axol robot adds Rigid Mode to ease two-arm teleop data collection for rigid objects — ycombinator · 2026-09-11