All-St3R SLAM: pixel-wise motion masking brings real-time dynamic SLAM to 10 FPS on RGB only

zhenjun_zhao · x · 2026-10-17

• Problem: Dynamic-scene SLAM suffers from over-removal (discarding temporarily static people/objects) and high compute cost.
• Method: Combines ego-motion pixel motion predicted by DUSt3R with point tracking from All-Tracker; by analyzing discrepancies between the two motion sources, it masks only actually-moving pixels — no predefined labels or segmentation models — so temporarily static parts remain usable for tracking.
• Performance: Runs at 10 FPS, handling standard 30 FPS video by subsampling every 3 frames; a big jump over SOTA dense dynamic SLAM (0.5–3 FPS) using RGB only.
• Accuracy: 2.46cm average ATE on Bonn RGB-D Dynamic and 2.27cm on TUM RGB-D Dynamic, on par with the best methods; ablations confirm both All-Tracker and confidence reweighting matter. Accepted to IROS 2026.

Original post →

More from Embodied

• Contact-UAN: MIT's residual world model fixes humanoid sim-to-real with 6.5 minutes of data — kevin_zakka · 2026-10-17
• ROS 2 LiDAR localization runs 390 m outdoors on handheld Livox MID-360 — rsasaki0109 · 2026-10-17
• Embodied Astra tops RoboMME at 92.1% — beating humans by just keeping full visual history in context — hankyang94 · 2026-10-17
• ROS weekly: ROSCon China kicks off in Shanghai, ROS By-The-Bay set for Oct 22 at Intrinsic — kscottz · 2026-10-17
• Simulated robotics eval: Claude Opus 5.5 preferred over Opus 5 in 14 of 16 comparisons — qinzytech · 2026-10-17
• FoundationVIO fuses IMU with foundation-model predictions to fix metric scale in feed-forward SLAM — zhenjun_zhao · 2026-10-17

Embodied channel →