<a href=\"https://cdn-uploads.huggingface.co/production/uploads/638a2bdb34cf0480e9abd6a4/4YgKoA7Xt_Son9nlw47wK.png\" rel=\"nofollow\"><img src=\"https://cdn-uploads.huggingface.co/production/uploads/638a2bdb34cf0480e9abd6a4/4YgKoA7Xt_Son9nlw47wK.png\" alt=\"hf\"></a></p>\n","updatedAt":"2026-09-11T06:57:33.020Z","author":{"_id":"638a2bdb34cf0480e9abd6a4","avatarUrl":"/avatars/70e84c5c188688db8b29455061a410a1.svg","fullname":"Zhiwen Yang","name":"upyzwup","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":0,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.3763089179992676},"editors":["upyzwup"],"editorAvatarUrls":["/avatars/70e84c5c188688db8b29455061a410a1.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2609.11156","authors":[{"_id":"6aa39d8047a406da7901e83c","name":"Zhiwen Yang","hidden":false},{"_id":"6aa39d8047a406da7901e83d","name":"Jiayin Li","hidden":false},{"_id":"6aa39d8047a406da7901e83e","name":"Chengyu Liu","hidden":false},{"_id":"6aa39d8047a406da7901e83f","name":"Hui Zhang","hidden":false},{"_id":"6aa39d8047a406da7901e840","name":"Bingzheng Wei","hidden":false},{"_id":"6aa39d8047a406da7901e841","name":"Yan Xu","hidden":false}],"publishedAt":"2026-09-10T00:00:00.000Z","submittedOnDailyAt":"2026-09-11T00:00:00.000Z","title":"UniH^3: Unifying Hierarchical Homogeneity and Heterogeneity for All-in-One Medical Image Restoration","submittedOnDailyBy":{"_id":"638a2bdb34cf0480e9abd6a4","avatarUrl":"/avatars/70e84c5c188688db8b29455061a410a1.svg","isPro":false,"fullname":"Zhiwen Yang","user":"upyzwup","type":"user","name":"upyzwup"},"summary":"All-in-One medical image restoration (MedIR) aims to address diverse tasks across modalities and degradation types using a single universal model. Existing methods typically prioritize modeling inter-task heterogeneity (e.g., distinct data distributions and degradation types). However, they largely neglect the inherent homogeneity present in medical images, such as widely shared anatomical structures within and across modalities, which can be leveraged to ease model training and improve generalization. To this end, we propose UniH3, a novel framework that Unifies Hierarchical Homogeneity and Heterogeneity for all-in-one medical image restoration. Specifically, to comprehensively exploit homogeneity, we introduce a Hierarchical Homogeneity Memory (H2M) module that progressively distills intra- and inter-task homogeneity priors from high-quality images during training, and adaptively retrieves the most relevant priors tailored to the input for guided restoration. These retrieved priors are then injected into the restoration pipeline via an efficient Homogeneity-Guided Attention (HGA) mechanism. Furthermore, to comprehensively address heterogeneity, we design a Hierarchical Heterogeneity Balancer (H2B) that mitigates both inter- and intra-task conflicts during optimization, facilitating balanced and effective multi-task learning. Extensive experiments on two large-scale benchmarks, MedIR-2D-500K and MedIR-3D-3K, demonstrate that UniH3 achieves state-of-the-art performance on both all-in-one and single-task medical image restoration. We hope this work establishes a strong benchmark and advances the development of general-purpose medical image restoration models. Code is available at https://github.com/Yaziwel/UniH3.","upvotes":2,"discussionId":"6aa39d8047a406da7901e842","projectPage":"https://github.com/Yaziwel/UniH3","githubRepo":"https://github.com/Yaziwel/UniH3","githubRepoAddedBy":"user","ai_summary":"UniH3 unifies hierarchical homogeneity and heterogeneity for universal medical image restoration via memory-based homogeneity priors and a heterogeneity balancer.","ai_keywords":["Hierarchical Homogeneity Memory","H2M","Homogeneity-Guided Attention","HGA","Hierarchical Heterogeneity Balancer","H2B","all-in-one medical image restoration","MedIR"],"ai_summary_model":"thinkingmachines/Inkling-Small","githubStars":5,"organization":{"_id":"63ba7720fc454697637969f1","name":"Beihang","fullname":"Beihang University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/63ba7666c138c8f2b7844b58/n98lZU9VWxYgWIkzE_6o4.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"638a2bdb34cf0480e9abd6a4","avatarUrl":"/avatars/70e84c5c188688db8b29455061a410a1.svg","isPro":false,"fullname":"Zhiwen Yang","user":"upyzwup","type":"user"},{"_id":"6a2da6c8ca070ee12c6e396c","avatarUrl":"/avatars/0355287dcabaa67dbc7f0b10b87451f9.svg","isPro":false,"fullname":"Joe Mama","user":"JoeMama123123123","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"63ba7720fc454697637969f1","name":"Beihang","fullname":"Beihang University","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/63ba7666c138c8f2b7844b58/n98lZU9VWxYgWIkzE_6o4.jpeg"},"query":{}}">
UniH^3: Unifying Hierarchical Homogeneity and Heterogeneity for All-in-One Medical Image Restoration
Abstract
UniH3 unifies hierarchical homogeneity and heterogeneity for universal medical image restoration via memory-based homogeneity priors and a heterogeneity balancer.
All-in-One medical image restoration (MedIR) aims to address diverse tasks across modalities and degradation types using a single universal model. Existing methods typically prioritize modeling inter-task heterogeneity (e.g., distinct data distributions and degradation types). However, they largely neglect the inherent homogeneity present in medical images, such as widely shared anatomical structures within and across modalities, which can be leveraged to ease model training and improve generalization. To this end, we propose UniH3, a novel framework that Unifies Hierarchical Homogeneity and Heterogeneity for all-in-one medical image restoration. Specifically, to comprehensively exploit homogeneity, we introduce a Hierarchical Homogeneity Memory (H2M) module that progressively distills intra- and inter-task homogeneity priors from high-quality images during training, and adaptively retrieves the most relevant priors tailored to the input for guided restoration. These retrieved priors are then injected into the restoration pipeline via an efficient Homogeneity-Guided Attention (HGA) mechanism. Furthermore, to comprehensively address heterogeneity, we design a Hierarchical Heterogeneity Balancer (H2B) that mitigates both inter- and intra-task conflicts during optimization, facilitating balanced and effective multi-task learning. Extensive experiments on two large-scale benchmarks, MedIR-2D-500K and MedIR-3D-3K, demonstrate that UniH3 achieves state-of-the-art performance on both all-in-one and single-task medical image restoration. We hope this work establishes a strong benchmark and advances the development of general-purpose medical image restoration models. Code is available at https://github.com/Yaziwel/UniH3.
Community
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2609.11156 in a model README.md to link it from this page.
Cite arxiv.org/abs/2609.11156 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2609.11156 in a Space README.md 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.