Metraipolis › archive › dc_193bf5535f77
Supervising the Path to Fine Scales: GalerkinFlow for Scientific-Field and Image Super-Resolution — arXiv:2608.16546
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.16546
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T19:44:27.145Z
- SHA-256 of the archived text
a1400bb0cf6f251d69835b66cea8f2b8842012867fd801ce090424afcdbf7b78- Stable URL of this page
https://metraipolis.com/documento/dc_193bf5535f77
How to cite
Supervising the Path to Fine Scales: GalerkinFlow for Scientific-Field and Image Super-Resolution — arXiv:2608.16546 — arXiv (en), https://arxiv.org/abs/2608.16546 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_193bf5535f77 (Metraipolis archive dc_193bf5535f77, sha256 a1400bb0cf6f251d69835b66cea8f2b8842012867fd801ce090424afcdbf7b78)
BibTeX
@misc{dc193bf5535f77,
title = {Supervising the Path to Fine Scales: GalerkinFlow for Scientific-Field and Image Super-Resolution — arXiv:2608.16546},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.16546},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_193bf5535f77; sha256 a1400bb0cf6f251d69835b66cea8f2b8842012867fd801ce090424afcdbf7b78},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.16546. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Most super-resolution models learn from paired data by supervising only the final high-resolution output. This provides little control over how the prediction should evolve between the downsampled observation and its fine target. We introduce GalerkinFlow, an equation-agnostic framework that turns each coarse--fine pair into supervision along an entire reconstruction path. At a random sample of intermediate states on the reconstruction path, the model predicts the coarse-to-fine residual velocity and uses coarse-anchor point to define a pseudo-endpoint. We show that the reconstruction loss of this pseudo-endpoint is exactly related to the intermediate velocity loss through a known time-dependent weight. Consequently, every intermediate state contributes supervision toward the same fine target, rather than serving only as an internal step toward an endpoint loss. Because intermediate states already reveal part of the missing fine-scale structure, we additionally supervise the coarse endpoint used during one-step inference. A finite-difference objective further constrains local spatial variation. GalerkinFlow combines convolutional features with scale-conditioned Galerkin operator mixing and requires no governing equation or physical metadata. It achieves the lowest raw-space errors among the evaluated equation-agnostic baselines on Navier--Stokes and Darcy Flow, while remaining competitive on DIV2K. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.LG). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Computational Engineering, Finance, and Science (cs.CE); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2608.16546 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-18T19:02:44.317Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).