PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
Wei Dong, Han Zhou, Terry Ji, Guanhua Zhao, Shahab Asoodeh, Yulun Zhang, Guangtao Zhai, Jun Chen, Xiaohong Liu · 2026
Read PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow by Wei Dong, Han Zhou, Terry Ji, Guanhua Zhao, Shahab Asoodeh, Yulun Zhang, Guangtao Zhai, Jun Chen, Xiaohong Liu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-14045.txt.
Description
Dong and eight co-authors offer PVRF, one system to strip many kinds of bad weather from real photos. Mixed and unseen weather damage plus distortion-led training often leave over-smooth results. PVRF first gauges soft weather odds and low-level traits with a question-answering module built on frozen vision-language models. Those gauges steer restoration nets through attribute-modulated normalization and weather-weighted adapters, forming an anchor estimate. A velocity-constrained rectified flow then refines that anchor.
Catalog record
| CREATOR | Wei Dong, Han Zhou, Terry Ji, Guanhua Zhao, Shahab Asoodeh, Yulun Zhang, Guangtao Zhai, Jun Chen, Xiaohong Liu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36267 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.14045