Hugging Face Daily Papers · · 3 min read

ReFlowSET: Representation-Aligned Latent Flow Matching for SAR-to-EO Image Translation

Mirrored from Hugging Face Daily Papers for archival readability. Support the source by reading on the original site.

ReFlowSET, a representation-aligned latent flow-matching framework for SAR-to-EO image translation with code &amp; checkpoints for all benchmarked methods fully released!</p>\n","updatedAt":"2026-09-02T01:43:37.565Z","author":{"_id":"66cc163fdd40cac2975b66d6","avatarUrl":"/avatars/343a13c7048ae73c6fad0d02dfc9a94a.svg","fullname":"Jeonghyeok Do","name":"JeonghyeokDo","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8167448043823242},"editors":["JeonghyeokDo"],"editorAvatarUrls":["/avatars/343a13c7048ae73c6fad0d02dfc9a94a.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2609.00968","authors":[{"_id":"6a977d31fe3c2f89286c38c0","name":"Jeonghyeok Do","hidden":false},{"_id":"6a977d31fe3c2f89286c38c1","name":"Seungchul Lee","hidden":false},{"_id":"6a977d31fe3c2f89286c38c2","name":"Munchurl Kim","hidden":false}],"publishedAt":"2026-09-01T00:00:00.000Z","submittedOnDailyAt":"2026-09-02T00:00:00.000Z","title":"ReFlowSET: Representation-Aligned Latent Flow Matching for SAR-to-EO Image Translation","submittedOnDailyBy":{"_id":"66cc163fdd40cac2975b66d6","avatarUrl":"/avatars/343a13c7048ae73c6fad0d02dfc9a94a.svg","isPro":false,"fullname":"Jeonghyeok Do","user":"JeonghyeokDo","type":"user","name":"JeonghyeokDo"},"summary":"SAR-to-EO image translation aims to generate electro-optical (EO) imagery from synthetic aperture radar (SAR) observations. Existing latent diffusion approaches typically inherit a predetermined autoencoder, although reconstruction fidelity can vary substantially across codecs and modalities. Because the latent codec affects the round-trip preservation of both SAR conditions and EO targets, codec selection constitutes a fundamental design choice; nevertheless, existing methods largely rely on codecs pretrained on natural images. To remedy this, we introduce ReFlowSET, a conditional latent flow-matching framework that selects its codec through a joint SAR--EO reconstruction audit. Rather than inheriting a heavyweight pretrained generator, ReFlowSET trains a substantially smaller conditional DiT from scratch in the selected latent space, using dual-stream SAR conditioning followed by joint feature refinement. To provide semantic guidance for this from-scratch training, intermediate noisy-EO features are aligned with clean target-EO representations extracted by a frozen vision foundation model. This alignment is used only during training and introduces no additional inference cost. Experiments on QXS-SAROPT and SAR2Opt demonstrate state-of-the-art performance across diverse perceptual fidelity and distributional metrics. Code and pretrained weights are publicly available at https://github.com/KAIST-VICLab/ReFlowSET.","upvotes":2,"discussionId":"6a977d32fe3c2f89286c38c3","projectPage":"https://kaist-viclab.github.io/ReFlowSET_site/","githubRepo":"https://github.com/KAIST-VICLab/ReFlowSET","githubRepoAddedBy":"user","ai_summary":"ReFlowSET selects a latent codec via joint SAR-EO reconstruction and trains a small conditional DiT with dual-stream conditioning and frozen vision-model alignment for high-fidelity SAR-to-EO translation.","ai_keywords":["latent flow-matching","conditional DiT","dual-stream SAR conditioning","joint feature refinement","vision foundation model","noisy-EO feature alignment","codec selection","SAR-to-EO translation"],"ai_summary_model":"thinkingmachines/Inkling-Small","githubStars":2,"organization":{"_id":"635b304962fb2bc1b52c6291","name":"KAIST","fullname":"KAIST","avatar":"https://www.gravatar.com/avatar/eba12517b1eaa0552a14abf582540dbd?d=retro&size=100"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"66cc163fdd40cac2975b66d6","avatarUrl":"/avatars/343a13c7048ae73c6fad0d02dfc9a94a.svg","isPro":false,"fullname":"Jeonghyeok Do","user":"JeonghyeokDo","type":"user"},{"_id":"631c386bc73939ffc0716a37","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1662793811119-noauth.jpeg","isPro":false,"fullname":"SeongWan Kim","user":"idgmatrix","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"635b304962fb2bc1b52c6291","name":"KAIST","fullname":"KAIST","avatar":"https://www.gravatar.com/avatar/eba12517b1eaa0552a14abf582540dbd?d=retro&size=100"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2609/2609.00968.md","query":{}}">
Papers
arxiv:2609.00968

ReFlowSET: Representation-Aligned Latent Flow Matching for SAR-to-EO Image Translation

Published on Sep 1
· Submitted by
Jeonghyeok Do
on Sep 2
Authors:
,

Abstract

ReFlowSET selects a latent codec via joint SAR-EO reconstruction and trains a small conditional DiT with dual-stream conditioning and frozen vision-model alignment for high-fidelity SAR-to-EO translation.

SAR-to-EO image translation aims to generate electro-optical (EO) imagery from synthetic aperture radar (SAR) observations. Existing latent diffusion approaches typically inherit a predetermined autoencoder, although reconstruction fidelity can vary substantially across codecs and modalities. Because the latent codec affects the round-trip preservation of both SAR conditions and EO targets, codec selection constitutes a fundamental design choice; nevertheless, existing methods largely rely on codecs pretrained on natural images. To remedy this, we introduce ReFlowSET, a conditional latent flow-matching framework that selects its codec through a joint SAR--EO reconstruction audit. Rather than inheriting a heavyweight pretrained generator, ReFlowSET trains a substantially smaller conditional DiT from scratch in the selected latent space, using dual-stream SAR conditioning followed by joint feature refinement. To provide semantic guidance for this from-scratch training, intermediate noisy-EO features are aligned with clean target-EO representations extracted by a frozen vision foundation model. This alignment is used only during training and introduces no additional inference cost. Experiments on QXS-SAROPT and SAR2Opt demonstrate state-of-the-art performance across diverse perceptual fidelity and distributional metrics. Code and pretrained weights are publicly available at https://github.com/KAIST-VICLab/ReFlowSET.

Community

ReFlowSET, a representation-aligned latent flow-matching framework for SAR-to-EO image translation with code & checkpoints for all benchmarked methods fully released!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images

· Sign up or log in to comment

Get this paper in your agent:

hf papers read 2609.00968
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper

Datasets citing this paper

No dataset linking this paper

Cite arxiv.org/abs/2609.00968 in a dataset README.md to link it from this page.

Spaces citing this paper

No Space linking this paper

Cite arxiv.org/abs/2609.00968 in a Space README.md to link it from this page.

Collections including this paper

No Collection including this paper

Add this paper to a collection to link it from this page.

Discussion (0)

Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.

Sign in →

No comments yet. Sign in and be the first to say something.

More from Hugging Face Daily Papers