Summary
Surflo is a feed-forward computer vision model that compresses a variable number of unposed RGB views into K latent tokens forming a single global state, then decodes oriented 3D surface points using flow matching. A guidance term applied at inference time correlates nearby points by injecting photometric gradients into the sampling process. According to the authors, Surflo matches or surpasses existing feed-forward baselines while running roughly an order of magnitude faster than optimization-based reconstruction methods. It is described as the only feed-forward approach that combines a global latent representation with arbitrary-resolution decoding of 3D surface points. The paper is authored by Antoine Guédon, Shu Nakamura, Nicolas Dufour, Jiahui Lei, Ko Nishino, and Angjoo Kanazawa, and is available on arXiv (2606.13644). This makes Surflo notable for balancing speed, consistency across views, and flexible output resolution in 3D surface reconstruction from multi-view RGB images.
Paper Overview
- Field: Computer Vision (CV)
- Authors: Antoine Guédon, Shu Nakamura, Nicolas Dufour, Jiahui Lei, Ko Nishino, Angjoo Kanazawa
- Published: 2026-06-11
- arXiv: 2606.13644
Abstract
We introduce Surflo, which compresses variable unposed RGB views into K latent tokens (one global state) and decodes oriented 3D surface points via flow matching. An inference-time guidance term correlates nearby points by injecting photometric gradient. Surflo matches or surpasses feed-forward baselines, runs an order of magnitude faster than optimization-based methods, and is the only feed-forward approach combining global latent with arbitrary-resolution decoding.
Key Contributions
- Compresses a variable number of unposed RGB views into K latent tokens forming a single global state.
- Decodes oriented 3D surface points via flow matching.
- An inference-time guidance term injects photometric gradients to correlate nearby points during sampling.
- Matches or surpasses feed-forward baselines while being roughly an order of magnitude faster than optimization-based methods.
- The only feed-forward method combining a global latent with arbitrary-resolution decoding.
---
*Auto-collected on 2026-06-14*
This page is an English static mirror generated for search and AI citation.
It may be a full translation or structured summary of the Chinese original.
Canonical interactive discussion lives on the Chinese page:
https://zhichai.net/topic/177981288