{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,16]],"date-time":"2026-07-16T05:18:10Z","timestamp":1784179090943,"version":"3.55.0"},"reference-count":84,"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"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,10,19]]},"DOI":"10.1109\/iccv51701.2025.01935","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"20812-20823","source":"Crossref","is-referenced-by-count":2,"title":["Dynamic-VLM: Simple Dynamic Visual Token Compression for VideoLLM"],"prefix":"10.1109","author":[{"given":"Han","family":"Wang","sequence":"first","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuxiang","family":"Nie","sequence":"additional","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yongjie","family":"Ye","sequence":"additional","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yanjie","family":"Wang","sequence":"additional","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shuai","family":"Li","sequence":"additional","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Haiyang","family":"Yu","sequence":"additional","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jinghui","family":"Lu","sequence":"additional","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Can","family":"Huang","sequence":"additional","affiliation":[{"name":"Bytedance Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00175"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1145\/1866029.1866080"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01605"},{"key":"ref4","article-title":"Token merging: Your vit but faster","author":"Bolya","year":"2022","journal-title":"arXiv preprint"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP48485.2024.10446658"},{"key":"ref6","first-page":"190","article-title":"Collecting highly parallel data for paraphrase evaluation","volume-title":"Proceedings of the 49th annual meeting of the association for computational linguistics: human language technologies","author":"Chen","year":"2011"},{"key":"ref7","article-title":"Allava: Harnessing gpt4v-synthesized data for a lite vision-language model","author":"Hardy Chen","year":"2024","journal-title":"arXiv preprint"},{"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 preprint"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72643-9_22"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0614"},{"key":"ref11","article-title":"Videollama 2: Advancing spatialtemporal modeling and audio understanding in video-llms","author":"Cheng","year":"2024","journal-title":"arXiv preprint"},{"key":"ref12","article-title":"Vicuna","author":"Chiang","year":"2023","journal-title":"An open-source chatbot impressing gpt-4 with 90 * chatgpt quality"},{"key":"ref13","author":"Cui","journal-title":"Comprehensive multimodal annotations with gpt-4o"},{"key":"ref14","article-title":"Video-ccam: Enhancing video-language understanding with causal cross-attention masks for short and long videos","author":"Fei","year":"2024","journal-title":"arXiv preprint"},{"key":"ref15","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 preprint"},{"key":"ref16","article-title":"Vita: Towards open-source interactive omni multimodal 1lm","author":"Fu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.670"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1007\/s10579-020-09517-1"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.149"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01300"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00592"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2024.3517625"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-016-0981-7"},{"key":"ref25","article-title":"Llava-onevision: Easy visual task transfer","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref26","article-title":"Llava-next-interleave: Tackling multi-image, video, and 3d in large multimodal models","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1007\/s11432-024-4321-9"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02095"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02095"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.775"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72952-2_19"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.342"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.02520"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.02484"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.02484"},{"key":"ref36","author":"Liu","year":"2024","journal-title":"Llava-next: Improved reasoning, ocr, and world knowledge"},{"key":"ref37","article-title":"Kangaroo: A powerful video-language model supporting long-context video input","author":"Liu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01296"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72998-0_1"},{"key":"ref40","article-title":"Tempcompass: Do video ll ~ms really understand videos?","author":"Liu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref41","article-title":"Iconqa: A new benchmark for abstract diagram understanding and visual language reasoning","author":"Lu","year":"2021","journal-title":"arXiv preprint"},{"key":"ref42","article-title":"Valley: Video assistant with large language model enhanced ability","author":"Luo","year":"2023","journal-title":"arXiv preprint"},{"key":"ref43","article-title":"Vista-llama: Reliable video narrator via equal distance to visual tokens","author":"Ma","year":"2023","journal-title":"arXiv preprint"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.679"},{"key":"ref45","article-title":"Videogpt+: Integrating image and video encoders for enhanced video understanding","author":"Maaz","year":"2024","journal-title":"arXiv preprint"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2004"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00331"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.findings-acl.177"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/WACV48630.2021.00225"},{"key":"ref50","article-title":"Mmiu: Multimodal multi-image understanding for evaluating large vision-language models","author":"Meng","year":"2024","journal-title":"arXiv preprint"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/ICDAR.2019.00156"},{"key":"ref52","article-title":"OpenAI","volume-title":"Gpt-4o mini: advancing cost-efficient intelligence","year":"2024"},{"key":"ref53","article-title":"OpenAI","volume-title":"Hello gpt-4o","year":"2024a"},{"key":"ref54","article-title":"OpenAI","volume-title":"Gpt-4v(ision) system card","year":"2024b"},{"key":"ref55","article-title":"Perception test: A diagnostic benchmark for multimodal video models","volume":"36","author":"Patraucean","year":"2024","journal-title":"Neural Information Processing Systems"},{"key":"ref56","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"International conference on machine learning","author":"Radford","year":"2021"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20074-8_9"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.01725"},{"key":"ref59","article-title":"Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context","author":"Team","year":"2024","journal-title":"arXiv preprint"},{"key":"ref60","volume-title":"Qwen2.5: A party of foundation models","author":"Team","year":"2024"},{"key":"ref61","article-title":"Llama: Open and efficient foundation language models","author":"Touvron","year":"2023","journal-title":"arXiv preprint"},{"key":"ref62","article-title":"How many unicorns are in this image? a safety evaluation benchmark for vision 11 ms","author":"Tu","year":"2023","journal-title":"arXiv preprint"},{"key":"ref63","article-title":"Muirbench: A comprehensive benchmark for robust multi-image understanding","author":"Wang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72670-5_10"},{"key":"ref65","article-title":"Longllava: Scaling multi-modal 11 ms to 1000 images efficiently via hybrid architecture","author":"Wang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref66","article-title":"Internvid: A large-scale video-text dataset for multimodal understanding and generation","author":"Wang","year":"2023","journal-title":"arXiv preprint"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02408"},{"key":"ref68","article-title":"Freeva: Offline mllm as training-free video assistant","author":"Wenhao","year":"2024","journal-title":"arXiv preprint"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00965"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1145\/3123266.3123427"},{"key":"ref71","article-title":"Slowfast-llava: A strong training-free baseline for video large language models","author":"Xu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00498"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33019127"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01100"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-demo.49"},{"key":"ref76","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 preprint"},{"key":"ref77","article-title":"Long context transfer from language to vision","author":"Zhang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref78","article-title":"Long context transfer from language to vision","author":"Zhang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref79","article-title":"Llama-adapter: Efficient fine-tuning of language models with zero-init attention","author":"Zhang","year":"2023","journal-title":"arXiv preprint"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-91813-1_4"},{"key":"ref81","author":"Zhang","year":"2024","journal-title":"Llavanext: A strong zero-shot video understanding model"},{"key":"ref82","article-title":"Mlvu: A comprehensive benchmark for multi-task long video understanding","author":"Zhou","year":"2024","journal-title":"arXiv preprint"},{"key":"ref83","article-title":"Minigpt-4: Enhancing vision-language understanding with advanced large language models","author":"Zhu","year":"2023","journal-title":"arXiv preprint"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.540"}],"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\/11444720.pdf?arnumber=11444720","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T05:19:41Z","timestamp":1777612781000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11444720\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":84,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.01935","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}