Surflo: NeurIPS 2026 Oral turns 2-80 unposed RGB views into clean 3D meshes
rsasaki0109 · x · 2026-09-29
Surflo, a Consistent 3D Surface Flow Model with Global State, has its official implementation released and was accepted as a NeurIPS 2026 Oral.
- Takes a variable number (2, 10, 30, even 80) of unposed RGB views and encodes the scene into a fixed-size latent
- Decodes via flow matching into an arbitrary number of oriented surface points, yielding a clean mesh
- Exceeds pointmap feed-forward models in quality while running an order of magnitude faster than optimization-based methods like Gaussian Wrapping, which need hundreds of views
Related event: Surflo reconstructs 3D surfaces from few photos, hits NeurIPS 2026 Oral(2 posts)→
More from Research
- Statistician mocks frequentists for patching one flaw and ignoring the rest — RexDouglass · 2026-09-29
- Pre-registration is just year-old priors, statistician argues in Bayesian-frequentist spat — RexDouglass · 2026-09-29
- 300M-param image model trained for €5000 claims SD 1.5-level performance — incorporo · 2026-09-29
- Cheap verifiers match costly ones in LLM post-training, saving up to 99.7% of grading cost — iScienceLuvr · 2026-09-29
- Telescopic LM trains one model valid at every depth, cutting quality-budget area 43% — iScienceLuvr · 2026-09-29
- ROFT: fine-tuning only on self-explanations matches GRPO on SWE-bench without RL — iScienceLuvr · 2026-09-29