Accurately modeling soft boundaries, e.g., hair and defocus blur, is a fundamental challenge in stereo conversion due to the ambiguous blending of foreground and background. Existing depth models primarily predict single-layer depth, leading to ambiguity in depth correspondence at soft boundaries. While matting techniques can capture opacity for layered modeling, they often struggle in complex scenes with multiple targets and usually require user intervention. This paper introduces αDepth, a layered representation that decomposes soft boundaries for high-fidelity stereo conversion. Specifically, we first resolve mixed color and depth ambiguity by estimating layered color and depth values at soft boundaries. Considering complex multi-target scenes, we design a Circular Alpha Representation (CAR) that shifts the paradigm from global target extraction to local boundary decomposition. Unlike prior matting methods restricted to a single foreground/background, CAR enables efficient scene-level inference without manual guidance. Extensive evaluations demonstrate that αDepth achieves state-of-the-art performance in stereo conversion, eliminating background bleeding and structural distortions at soft boundaries.</p>\n","updatedAt":"2026-06-03T09:29:34.996Z","author":{"_id":"657dc1576dc01435cd9029d8","avatarUrl":"/avatars/3bba11ac7659fce61aeaedf40e2057a8.svg","fullname":"Xiang Zhang","name":"XiangZ","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":4,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8519003987312317},"editors":["XiangZ"],"editorAvatarUrls":["/avatars/3bba11ac7659fce61aeaedf40e2057a8.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.00386","authors":[{"_id":"6a1e68b7808ddbc3c7d43e5b","name":"Xiang Zhang","hidden":false},{"_id":"6a1e68b7808ddbc3c7d43e5c","name":"Yang Zhang","hidden":false},{"_id":"6a1e68b7808ddbc3c7d43e5d","name":"Lukas Mehl","hidden":false},{"_id":"6a1e68b7808ddbc3c7d43e5e","name":"Karlis Martins Briedis","hidden":false},{"_id":"6a1e68b7808ddbc3c7d43e5f","name":"Markus Gross","hidden":false},{"_id":"6a1e68b7808ddbc3c7d43e60","name":"Christopher Schroers","hidden":false}],"publishedAt":"2026-05-29T00:00:00.000Z","submittedOnDailyAt":"2026-06-03T00:00:00.000Z","title":"αDepth: Learning Single-Pass Soft Boundary Decomposition for Stereo Conversion","submittedOnDailyBy":{"_id":"657dc1576dc01435cd9029d8","avatarUrl":"/avatars/3bba11ac7659fce61aeaedf40e2057a8.svg","isPro":false,"fullname":"Xiang Zhang","user":"XiangZ","type":"user","name":"XiangZ"},"summary":"Accurately modeling soft boundaries, e.g., hair and defocus blur, is a fundamental challenge in stereo conversion due to the ambiguous blending of foreground and background. Existing depth models primarily predict single-layer depth, leading to ambiguity in depth correspondence at soft boundaries. While matting techniques can capture opacity for layered modeling, they often struggle in complex scenes with multiple targets and usually require user intervention. This paper introduces αDepth, a layered representation that decomposes soft boundaries for high-fidelity stereo conversion. Specifically, we first resolve mixed color and depth ambiguity by estimating layered color and depth values at soft boundaries. Considering complex multi-target scenes, we design a Circular Alpha Representation (CAR) that shifts the paradigm from global target extraction to local boundary decomposition. Unlike prior matting methods restricted to a single foreground/background, CAR enables efficient scene-level inference without manual guidance. Extensive evaluations demonstrate that αDepth achieves state-of-the-art performance in stereo conversion, eliminating background bleeding and structural distortions at soft boundaries.","upvotes":1,"discussionId":"6a1e68b7808ddbc3c7d43e61","ai_summary":"αDepth introduces a layered representation with Circular Alpha Representation (CAR) to address soft boundary challenges in stereo conversion through local boundary decomposition and efficient scene-level inference.","ai_keywords":["layered representation","soft boundaries","stereo conversion","depth modeling","matting techniques","Circular Alpha Representation","scene-level inference","background bleeding","structural distortions"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","organization":{"_id":"63263d7db8e57aab1a778773","name":"ethz","fullname":"ETH Zurich","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/5dd96eb166059660ed1ee413/xMcrQI8Yx8o697uhiCcoA.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"657dc1576dc01435cd9029d8","avatarUrl":"/avatars/3bba11ac7659fce61aeaedf40e2057a8.svg","isPro":false,"fullname":"Xiang Zhang","user":"XiangZ","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"63263d7db8e57aab1a778773","name":"ethz","fullname":"ETH Zurich","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/5dd96eb166059660ed1ee413/xMcrQI8Yx8o697uhiCcoA.jpeg"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.00386.md"}">
αDepth: Learning Single-Pass Soft Boundary Decomposition for Stereo Conversion
Abstract
αDepth introduces a layered representation with Circular Alpha Representation (CAR) to address soft boundary challenges in stereo conversion through local boundary decomposition and efficient scene-level inference.
Accurately modeling soft boundaries, e.g., hair and defocus blur, is a fundamental challenge in stereo conversion due to the ambiguous blending of foreground and background. Existing depth models primarily predict single-layer depth, leading to ambiguity in depth correspondence at soft boundaries. While matting techniques can capture opacity for layered modeling, they often struggle in complex scenes with multiple targets and usually require user intervention. This paper introduces αDepth, a layered representation that decomposes soft boundaries for high-fidelity stereo conversion. Specifically, we first resolve mixed color and depth ambiguity by estimating layered color and depth values at soft boundaries. Considering complex multi-target scenes, we design a Circular Alpha Representation (CAR) that shifts the paradigm from global target extraction to local boundary decomposition. Unlike prior matting methods restricted to a single foreground/background, CAR enables efficient scene-level inference without manual guidance. Extensive evaluations demonstrate that αDepth achieves state-of-the-art performance in stereo conversion, eliminating background bleeding and structural distortions at soft boundaries.
Community
Accurately modeling soft boundaries, e.g., hair and defocus blur, is a fundamental challenge in stereo conversion due to the ambiguous blending of foreground and background. Existing depth models primarily predict single-layer depth, leading to ambiguity in depth correspondence at soft boundaries. While matting techniques can capture opacity for layered modeling, they often struggle in complex scenes with multiple targets and usually require user intervention. This paper introduces αDepth, a layered representation that decomposes soft boundaries for high-fidelity stereo conversion. Specifically, we first resolve mixed color and depth ambiguity by estimating layered color and depth values at soft boundaries. Considering complex multi-target scenes, we design a Circular Alpha Representation (CAR) that shifts the paradigm from global target extraction to local boundary decomposition. Unlike prior matting methods restricted to a single foreground/background, CAR enables efficient scene-level inference without manual guidance. Extensive evaluations demonstrate that αDepth achieves state-of-the-art performance in stereo conversion, eliminating background bleeding and structural distortions at soft boundaries.
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/2606.00386 in a model README.md to link it from this page.
Cite arxiv.org/abs/2606.00386 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2606.00386 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.