LightSplat: real-time high-fidelity 3D Gaussian SLAM with loop closure at 8 FPS
zhenjun_zhao · x · 2026-09-11
LightSplat, accepted to IROS 2026, is a hybrid-representation RGB-D SLAM framework: local sparse features drive robust fast tracking while a dual-thread backend progressively builds dense Gaussian submaps; online loop closure via feature-accelerated 3DGS registration plus pose graph optimization refines map consistency. It reaches near-SOTA reconstruction quality on multiple datasets and a real robotic platform, handles practical camera motions, and maintains an average 8 FPS for real-world deployment.
More from Embodied
- Embodied AI startups need a different operational discipline, investor says — ___Mufasaa · 2026-10-27
- π0.5 study: raising demos from 10 to 438 lifts Place Ball in Bowl success from 7% to 83% — stingning · 2026-10-27
- GPT-6 Astra controls dexterous hands zero-shot at just 5.2% success in 500 trials — stingning · 2026-10-27
- HandsOn benchmark: 66 verified tasks and 18,000 human demos for dexterous manipulation — stingning · 2026-10-27
- HandsOn dexterous-manipulation benchmark: best policy 29.1%, GPT-6 Astra just 5.2% — stingning · 2026-10-27
- DexPort: a handheld interface with a real robot hand for dexterous-hand data collection — pabbeel · 2026-10-27