CLoSeR adds loop closure to streaming reconstruction for drift-free km-scale 3D
zhenjun_zhao · x · 2026-10-02
New paper CLoSeR (Moyang Li, Zihan Zhu et al.) tackles tracking drift from error accumulation in feedforward streaming reconstruction foundation models by reintroducing loop closure:
- Detects loop candidates via global descriptor retrieval and builds loop-conditioned windows for relative pose estimation
- Observing that the streaming backbone yields globally consistent scale, optimizes all frame poses on the SE(3) manifold, avoiding Sim(3)/SL(4) pose graph optimization used in prior work
- Reduces drift and produces consistent geometry on kilometer-scale sequences, significantly outperforming SOTA; code released
More from Research
- Symbolic supervision gives video models equal reasoning at ~1/30 the compute, Rubik's Cube study finds — niloofar_mire · 2026-10-03
- Sony AI open-sources PAVAS, a physics-aware video-to-audio model accepted as CVPR 2026 Oral — mittu1204 · 2026-10-03
- Maryland RCT: GPT-4o tutor offering left students 4 points lower, 15% used it — lmoroney · 2026-10-03
- OpenAI paper: simulating deployment predicts LLM safety before release — JacquesThibs · 2026-10-03
- PyTorch Helion DSL integrated into vLLM beats CUTLASS and DeepGEMM with 10%+ throughput gains — PyTorch · 2026-10-03
- Neural memory as an unbounded 2D lattice: writes and reads through a tiny 3x3 bump — CFGeek · 2026-10-03