Slot3R: set-associative spatial memory cuts streaming 3D reconstruction errors by up to 72%
zhenjun_zhao · x · 2026-10-09
Slot3R retrofits Point3R with a training-free, set-associative spatial memory: location determines address only, not whether observations merge, with a bounded sparse readout decoupling storage from per-frame access. It cuts point-cloud accuracy errors by 57.1–63.1% on 7Scenes and 64.0–72.0% on NeuralRGBD, lowers ATE on all three pose benchmarks, and runs 19 FPS up to 1000 frames where baselines OOM at 800.
More from Embodied
- MIT CSAIL's X-Hinges 3D-Prints Everyday Objects With Built-In Sensing Across Three Motion Axes — MIT News AI · 2026-10-23
- Formic raises $50M Series B to scale physical AI across 100+ US factories — garrytan · 2026-10-23
- fal starts fine-tuning a physical robot to draw, results next week — OdinLovis · 2026-10-23
- PEARL demo: an LLM agent that checks experiment results, revises plans, and re-tests — KevinKaichuang · 2026-10-23
- CU Boulder's ATLAS Institute Opens Tenure-Track Robotics Professorship — ryosuzk · 2026-10-23
- MiTaS multi-resolution tactile imitation learning hits 80% success vs 35.5% vision-only baseline — Jan_R_Peters · 2026-10-23