https://rookiexiong7.github.io/projects/SeC/
💻 Code: https://github.com/OpenIXCLab/SeC
📊 Dataset: https://huggingface.co/datasets/OpenIXCLab/SeCVOS

\n","updatedAt":"2025-07-22T05:54:25.637Z","author":{"_id":"64b4eec4faa3181a5eab9c46","avatarUrl":"/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg","fullname":"Jiaqi Wang","name":"myownskyW7","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":25,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7187488675117493},"editors":["myownskyW7"],"editorAvatarUrls":["/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg"],"reactions":[{"reaction":"🔥","users":["AdinaY"],"count":1}],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2507.15852","authors":[{"_id":"687efbab33947f780d9b4aee","user":{"_id":"65ab5332043d53781a115475","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65ab5332043d53781a115475/UaxSFDWteYsByzx7G_KKy.jpeg","isPro":false,"fullname":"Zhixiong Zhang (SII)","user":"rookiexiong","type":"user"},"name":"Zhixiong Zhang","status":"claimed_verified","statusLastChangedAt":"2025-07-22T07:48:03.855Z","hidden":false},{"_id":"687efbab33947f780d9b4aef","name":"Shuangrui Ding","hidden":false},{"_id":"687efbab33947f780d9b4af0","user":{"_id":"68943a6e8d3fb6db77ce2874","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/7LBz6WhQmiKxOqZByJhr9.jpeg","isPro":false,"fullname":"Xiaoyi Dong","user":"LightDong","type":"user"},"name":"Xiaoyi Dong","status":"claimed_verified","statusLastChangedAt":"2025-08-07T10:42:58.628Z","hidden":false},{"_id":"687efbab33947f780d9b4af1","name":"Songxin He","hidden":false},{"_id":"687efbab33947f780d9b4af2","name":"Jianfan Lin","hidden":false},{"_id":"687efbab33947f780d9b4af3","name":"Junsong Tang","hidden":false},{"_id":"687efbab33947f780d9b4af4","user":{"_id":"63859cf3b2906edaf83af9f0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63859cf3b2906edaf83af9f0/kajwuVzd4pDucSPlwghxo.png","isPro":true,"fullname":"Yuhang Zang","user":"yuhangzang","type":"user"},"name":"Yuhang Zang","status":"claimed_verified","statusLastChangedAt":"2025-07-22T13:05:45.709Z","hidden":false},{"_id":"687efbab33947f780d9b4af5","name":"Yuhang Cao","hidden":false},{"_id":"687efbab33947f780d9b4af6","name":"Dahua Lin","hidden":false},{"_id":"687efbab33947f780d9b4af7","user":{"_id":"64b4eec4faa3181a5eab9c46","avatarUrl":"/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg","isPro":true,"fullname":"Jiaqi Wang","user":"myownskyW7","type":"user"},"name":"Jiaqi Wang","status":"claimed_verified","statusLastChangedAt":"2025-08-14T13:40:31.337Z","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/64b4eec4faa3181a5eab9c46/pgY_J9jZ7-tdgqX8uLYeA.mp4"],"publishedAt":"2025-07-21T17:59:02.000Z","submittedOnDailyAt":"2025-07-22T04:24:25.629Z","title":"SeC: Advancing Complex Video Object Segmentation via Progressive Concept\n Construction","submittedOnDailyBy":{"_id":"64b4eec4faa3181a5eab9c46","avatarUrl":"/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg","isPro":true,"fullname":"Jiaqi Wang","user":"myownskyW7","type":"user"},"summary":"Video Object Segmentation (VOS) is a core task in computer vision, requiring\nmodels to track and segment target objects across video frames. Despite notable\nadvances with recent efforts, current techniques still lag behind human\ncapabilities in handling drastic visual variations, occlusions, and complex\nscene changes. This limitation arises from their reliance on appearance\nmatching, neglecting the human-like conceptual understanding of objects that\nenables robust identification across temporal dynamics. Motivated by this gap,\nwe propose Segment Concept (SeC), a concept-driven segmentation framework that\nshifts from conventional feature matching to the progressive construction and\nutilization of high-level, object-centric representations. SeC employs Large\nVision-Language Models (LVLMs) to integrate visual cues across diverse frames,\nconstructing robust conceptual priors. During inference, SeC forms a\ncomprehensive semantic representation of the target based on processed frames,\nrealizing robust segmentation of follow-up frames. Furthermore, SeC adaptively\nbalances LVLM-based semantic reasoning with enhanced feature matching,\ndynamically adjusting computational efforts based on scene complexity. To\nrigorously assess VOS methods in scenarios demanding high-level conceptual\nreasoning and robust semantic understanding, we introduce the Semantic Complex\nScenarios Video Object Segmentation benchmark (SeCVOS). SeCVOS comprises 160\nmanually annotated multi-scenario videos designed to challenge models with\nsubstantial appearance variations and dynamic scene transformations. In\nparticular, SeC achieves an 11.8-point improvement over SAM 2.1 on SeCVOS,\nestablishing a new state-of-the-art in concept-aware video object segmentation.","upvotes":38,"discussionId":"687efbab33947f780d9b4af8","projectPage":"https://rookiexiong7.github.io/projects/SeC","githubRepo":"https://github.com/OpenIXCLab/SeC","githubRepoAddedBy":"user","ai_summary":"A concept-driven segmentation framework using Large Vision-Language Models improves video object segmentation by integrating high-level semantic reasoning and adapting to scene complexity.","ai_keywords":["Video Object Segmentation","VOS","Large Vision-Language Models","LVLMs","semantic representation","feature matching","SeCVOS","concept-aware segmentation"],"githubStars":269},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"64b4eec4faa3181a5eab9c46","avatarUrl":"/avatars/bcc9bf5cbf67546ad2b4c9ec8b96ac96.svg","isPro":true,"fullname":"Jiaqi Wang","user":"myownskyW7","type":"user"},{"_id":"6564a63de8d6a163c0a4fd2b","avatarUrl":"/avatars/a4b7cc9ea6e4023f150fd745cab3a883.svg","isPro":false,"fullname":"Azusa Aisaka","user":"Bujiazi","type":"user"},{"_id":"64adfeac4beffa272dfaef21","avatarUrl":"/avatars/883f6ba38b993476115dfafcef9ce3c1.svg","isPro":false,"fullname":"Yifei Li","user":"JoeLeelyf","type":"user"},{"_id":"65000bef18830fabea469fdd","avatarUrl":"/avatars/b320c77dfad039d9f9c54127f610d44f.svg","isPro":false,"fullname":"Cao Yuhang","user":"yhcao","type":"user"},{"_id":"64b51fd8bcfd8542d6473d9a","avatarUrl":"/avatars/ceaa73b79f448996187f07733d96b800.svg","isPro":false,"fullname":"yujie","user":"yujieouo","type":"user"},{"_id":"6444f0a8b272430bdbf11785","avatarUrl":"/avatars/5135f817e638e97b280a28ba90d4381c.svg","isPro":false,"fullname":"laolao","user":"laolao77","type":"user"},{"_id":"64de20c5808492ba6e65d124","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64de20c5808492ba6e65d124/58IX_TI5vJw73qS1knw56.jpeg","isPro":false,"fullname":"Zhang Mengchen","user":"Dubhe-zmc","type":"user"},{"_id":"63fda3fced9eead590ff6918","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1677566802735-noauth.jpeg","isPro":false,"fullname":"Zeyi Sun","user":"Zery","type":"user"},{"_id":"667a5d6669c3e097d57356f6","avatarUrl":"/avatars/630a3bf4c06e8e87c37b829831912b39.svg","isPro":false,"fullname":"Rui","user":"srding","type":"user"},{"_id":"65ab5332043d53781a115475","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65ab5332043d53781a115475/UaxSFDWteYsByzx7G_KKy.jpeg","isPro":false,"fullname":"Zhixiong Zhang (SII)","user":"rookiexiong","type":"user"},{"_id":"63859cf3b2906edaf83af9f0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63859cf3b2906edaf83af9f0/kajwuVzd4pDucSPlwghxo.png","isPro":true,"fullname":"Yuhang Zang","user":"yuhangzang","type":"user"},{"_id":"6373037cd3ba6bd3f9bc32fa","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6373037cd3ba6bd3f9bc32fa/dIL4AIcL24QL_Rp4B9ewS.png","isPro":false,"fullname":"Shuai Yang","user":"ysmikey","type":"user"}],"acceptLanguages":["*"],"dailyPaperRank":0}">