arXiv Open Access 2025

Saving Foundation Flow-Matching Priors for Inverse Problems

Yuxiang Wan Ryan Devera Wenjie Zhang Ju Sun

Lihat Sumber

Abstrak

Foundation flow-matching (FM) models promise a universal prior for solving inverse problems (IPs), yet today they trail behind domain-specific or even untrained priors. How can we unlock their potential? We introduce FMPlug, a plug-in framework that redefines how foundation FMs are used in IPs. FMPlug combines an instance-guided, time-dependent warm-start strategy with a sharp Gaussianity regularization, adding problem-specific guidance while preserving the Gaussian structures. This leads to a significant performance boost across image restoration and scientific IPs. Our results point to a path for making foundation FM models practical, reusable priors for IP solving.

Topik & Kata Kunci

cs.LG cs.CV eess.IV eess.SP

Penulis (4)

Yuxiang Wan

Ryan Devera

Wenjie Zhang

Ju Sun

Format Sitasi

APA MLA BibTeX

Wan, Y., Devera, R., Zhang, W., Sun, J. (2025). Saving Foundation Flow-Matching Priors for Inverse Problems. https://arxiv.org/abs/2511.16520

Akses Cepat

Lihat di Sumber

Informasi Jurnal

Tahun Terbit: 2025
Bahasa: en
Sumber Database: arXiv
Akses: Open Access ✓