{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T10:13:20Z","timestamp":1777889600877,"version":"3.51.4"},"reference-count":86,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,10,19]],"date-time":"2025-10-19T00:00:00Z","timestamp":1760832000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,10,19]],"date-time":"2025-10-19T00:00:00Z","timestamp":1760832000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62422606,62201484"],"award-info":[{"award-number":["62422606,62201484"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100012166","name":"National Key R&D Program of China","doi-asserted-by":"publisher","award":["2022ZD0160101"],"award-info":[{"award-number":["2022ZD0160101"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,10,19]]},"DOI":"10.1109\/iccv51701.2025.02016","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"21710-21720","source":"Crossref","is-referenced-by-count":0,"title":["DisCo: Towards Distinct and Coherent Visual Encapsulation in Video MLLMs"],"prefix":"10.1109","author":[{"given":"Jiahe","family":"Zhao","sequence":"first","affiliation":[{"name":"The University of Hong Kong"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Rongkun","family":"Zheng","sequence":"additional","affiliation":[{"name":"The University of Hong Kong"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yi","family":"Wang","sequence":"additional","affiliation":[{"name":"Shanghai Artificial Intelligence Laboratory"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Helin","family":"Wang","sequence":"additional","affiliation":[{"name":"Fudan University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hengshuang","family":"Zhao","sequence":"additional","affiliation":[{"name":"The University of Hong Kong"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Gpt-4 technical report","author":"Achiam","year":"2023","journal-title":"arXiv:2303.08774"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref3","article-title":"Qwen-vl: A versatile vision-language model for understanding, localization, text reading, and beyond","author":"Bai","year":"2023","journal-title":"arXiv:2308.12966"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00175"},{"key":"ref5","article-title":"Language models are few-shot learners","volume-title":"NeurIPS","author":"Brown","year":"2020"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58452-8_13"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01311"},{"key":"ref8","article-title":"Minigpt-v2: large language model as a unified interface for vision-language multi-task learning","author":"Chen","year":"2023","journal-title":"arXiv:2310.09478"},{"key":"ref9","article-title":"Sharegpt4v: Improving large multi-modal models with better captions","author":"Chen","year":"2023","journal-title":"arXiv:2311.12793"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0614"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00630"},{"key":"ref12","article-title":"Egoplanbench: Benchmarking egocentric embodied planning with multimodal large language models","author":"Chen","year":"2023","journal-title":"arXiv:2312.06722"},{"key":"ref13","article-title":"Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks","volume-title":"CVPR","author":"Chen","year":"2024"},{"key":"ref14","article-title":"Videollama 2: Advancing spatialtemporal modeling and audio understanding in video-llms","author":"Cheng","year":"2024","journal-title":"arXiv:2406.07476"},{"key":"ref15","author":"Chiang","year":"2023","journal-title":"Vicuna: An open-source chatbot impressing gpt-4 with 90\\%* chatgpt quality"},{"key":"ref16","article-title":"Mobilevlm v2: Faster and stronger baseline for vision language model","author":"Chu","year":"2024","journal-title":"arXiv:2402.03766"},{"key":"ref17","article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018","journal-title":"arXiv:1810.04805"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2010.11929"},{"key":"ref19","article-title":"Palm-e: An embodied multimodal language model","author":"Driess","year":"2023","journal-title":"arXiv:2303.03378"},{"key":"ref20","article-title":"The llama 3 herd of models","author":"Dubey","year":"2024","journal-title":"arXiv:2407.21783"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01855"},{"key":"ref22","article-title":"Video-mme: The first-ever comprehensive evaluation benchmark of multi-modal 11 ms in video analysis","author":"Fu","year":"2024","journal-title":"arXiv:2405.21075"},{"key":"ref23","article-title":"Planting a seed of vision in large language model","author":"Ge","year":"2023","journal-title":"arXiv:2307.08041"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.622"},{"key":"ref25","article-title":"Lora: Low-rank adaptation of large language models","volume-title":"ICLR","author":"Hu","year":"2022"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"ref27","article-title":"Mistral 7b","author":"Jiang","year":"2023","journal-title":"arXiv:2310.06825"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01300"},{"key":"ref29","article-title":"The kinetics human action video dataset","author":"Kay","year":"2017","journal-title":"arXiv:1705.06950"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73016-0_16"},{"key":"ref31","article-title":"Mimic-it: Multi-modal in-context instruction tuning","author":"Li","year":"2023","journal-title":"arXiv:2306.05425"},{"key":"ref32","article-title":"Otter: a multi-modal model with in-context instruction tuning","author":"Li","year":"2023","journal-title":"arXiv:2305.03726"},{"key":"ref33","article-title":"Llava-onevision: Easy visual task transfer","author":"Li","year":"2024","journal-title":"arXiv:2408.03326"},{"key":"ref34","article-title":"Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation","volume-title":"ICML","author":"Li","year":"2022"},{"key":"ref35","article-title":"Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models","volume-title":"ICML","author":"Li","year":"2023"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1007\/s11432-024-4321-9"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02095"},{"key":"ref38","article-title":"Topa: Extend large language models for video understanding via text-only pre-alignment","author":"Li","year":"2024","journal-title":"arXiv:2405.13911"},{"key":"ref39","article-title":"Tokenpacker: Efficient visual projector for multimodal 11 m","author":"Li","year":"2024","journal-title":"arXiv:2407.02392"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72952-2_19"},{"key":"ref41","article-title":"Video-llava: Learning united visual representation by alignment before projection","author":"Lin","year":"2023","journal-title":"arXiv:2311.10122"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02484"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72998-0_1"},{"key":"ref46","article-title":"Objectcentric learning with slot attention","volume-title":"NeurIPS","author":"Locatello","year":"2020"},{"key":"ref47","article-title":"The flan collection: Designing data and methods for effective instruction tuning","volume-title":"ICML","author":"Longpre","year":"2023"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.679"},{"key":"ref49","article-title":"Videogpt+: Integrating image and video encoders for enhanced video understanding","author":"Maaz","year":"2024","journal-title":"arXiv:2406.09418"},{"key":"ref50","article-title":"Egoschema: A diagnostic benchmark for very longform video language understanding","volume-title":"NeurIPS","author":"Mangalam","year":"2023"},{"key":"ref51","article-title":"Embodiedgpt: Vision-language pre-training via embodied chain of thought","volume-title":"NeurIPS","author":"Mu","year":"2024"},{"key":"ref52","article-title":"Perception test: A diagnostic benchmark for multimodal video models","volume-title":"NeurIPS","author":"Patraucean","year":"2024"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.303"},{"key":"ref54","article-title":"Streaming long video understanding with large language models","author":"Qian","year":"2024","journal-title":"arXiv:2405.16009"},{"key":"ref55","article-title":"Learning transferable visual models from natural language supervision","volume-title":"ICML","author":"Radford","year":"2021"},{"key":"ref56","article-title":"Attention is all you need","volume-title":"NeurIPS","author":"Vaswani","year":"2017"},{"key":"ref57","article-title":"Qwen2-vl: Enhancing vision-language model\u2019s perception of the world at any resolution","author":"Wang","year":"2024","journal-title":"arXiv:2409.12191"},{"key":"ref58","article-title":"Visionllm: Large language model is also an openended decoder for vision-centric tasks","volume-title":"NeurIPS","author":"Wang","year":"2024"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.findings-emnlp.1168"},{"key":"ref60","article-title":"Internvideo2: Scaling video foundation models for multimodal video understanding","volume-title":"ECCV","author":"Wang","year":"2024"},{"key":"ref61","article-title":"Star: A benchmark for situated reasoning in real-world videos","volume-title":"NeurIPS","author":"Wu","year":"2021"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00965"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.571"},{"key":"ref64","article-title":"Slot-vlm: Slowfast slots for video-language modeling","author":"Xu","year":"2024","journal-title":"arXiv:2402.13088"},{"key":"ref65","article-title":"Pllava: Parameter-free llava extension from images to videos for video dense captioning","author":"Xu","year":"2024","journal-title":"arXiv:2404.16994"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1109\/LRA.2024.3440097"},{"key":"ref67","article-title":"Deco: Decoupling token compression from semantic abstraction in multimodal large language models","author":"Yao","year":"2024","journal-title":"arXiv:2405.20985"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01239"},{"key":"ref69","article-title":"Clevrer: Collision events for video representation and reasoning","volume-title":"ICLR","author":"Yi","year":"2020"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1036\/1097-8542.253500"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46475-6_5"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02221"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-demo.49"},{"key":"ref74","article-title":"Internlm-xcomposer-2.5: A versatile large vision language model supporting long-contextual input and output","author":"Zhang","year":"2024","journal-title":"arXiv:2407.03320"},{"key":"ref75","article-title":"Direct preference optimization of video large multimodal models from language model reward","author":"Zhang","year":"2024","journal-title":"arXiv:2404.01258"},{"key":"ref76","article-title":"Llava-mini: Efficient image and video large multimodal models with one vision token","volume-title":"ICLR","author":"Zhang","year":"2025"},{"key":"ref77","author":"Zhang","year":"2024","journal-title":"Llavanext: A strong zero-shot video understanding model"},{"key":"ref78","article-title":"Mmicl: Empowering vision-language model with multi-modal in-context learning","author":"Zhao","year":"2023","journal-title":"arXiv:2309.07915"},{"key":"ref79","article-title":"Antgpt: Can large language models help long-term action anticipation from videos?","volume-title":"ICLR","author":"Zhao","year":"2024"},{"key":"ref80","article-title":"Judging llm-as-a-judge with mt-bench and chatbot arena","volume-title":"NeurIPS","author":"Zheng","year":"2023"},{"key":"ref81","article-title":"Syncvis: Synchronized video instance segmentation","volume-title":"NeurIPS","author":"Zheng","year":"2024"},{"key":"ref82","article-title":"Villa: Video reasoning segmentation with large language model","author":"Zheng","year":"2024","journal-title":"arXiv:2407.14500"},{"key":"ref83","article-title":"Mlvu: A comprehensive benchmark for multi-task long video understanding","author":"Zhou","year":"2024","journal-title":"arXiv:2406.04264"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v32i1.12342"},{"key":"ref85","article-title":"Minigpt-4: Enhancing visionlanguage understanding with advanced large language models","author":"Zhu","year":"2023","journal-title":"arXiv:2304.10592"},{"key":"ref86","article-title":"Vl-gpt: A generative pre-trained transformer for vision and language understanding and generation","author":"Zhu","year":"2023","journal-title":"arXiv:2312.09251"}],"event":{"name":"2025 IEEE\/CVF International Conference on Computer Vision (ICCV)","location":"Honolulu, HI, USA","start":{"date-parts":[[2025,10,19]]},"end":{"date-parts":[[2025,10,25]]}},"container-title":["2025 IEEE\/CVF International Conference on Computer Vision (ICCV)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11443115\/11443287\/11445473.pdf?arnumber=11445473","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T05:13:31Z","timestamp":1777612411000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11445473\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":86,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.02016","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}