{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T16:25:15Z","timestamp":1778171115542,"version":"3.51.4"},"reference-count":111,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"4","license":[{"start":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T00:00:00Z","timestamp":1775001600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T00:00:00Z","timestamp":1775001600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T00:00:00Z","timestamp":1775001600000},"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":"crossref","award":["62171281"],"award-info":[{"award-number":["62171281"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100003399","name":"Science and Technology Commission of Shanghai Municipality","doi-asserted-by":"crossref","award":["20DZ1200203"],"award-info":[{"award-number":["20DZ1200203"]}],"id":[{"id":"10.13039\/501100003399","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100003399","name":"Science and Technology Commission of Shanghai Municipality","doi-asserted-by":"crossref","award":["2021SHZDZX0102"],"award-info":[{"award-number":["2021SHZDZX0102"]}],"id":[{"id":"10.13039\/501100003399","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Circuits Syst. Video Technol."],"published-print":{"date-parts":[[2026,4]]},"DOI":"10.1109\/tcsvt.2025.3643469","type":"journal-article","created":{"date-parts":[[2025,12,12]],"date-time":"2025-12-12T18:37:50Z","timestamp":1765564670000},"page":"4464-4478","source":"Crossref","is-referenced-by-count":5,"title":["MG-LLaVA: Toward Multi-Granularity Visual Instruction Tuning"],"prefix":"10.1109","volume":"36","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-1313-9451","authenticated-orcid":false,"given":"Xiangyu","family":"Zhao","sequence":"first","affiliation":[{"name":"Shanghai Jiao Tong University, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0550-8247","authenticated-orcid":false,"given":"Xiangtai","family":"Li","sequence":"additional","affiliation":[{"name":"Shanghai AI Laboratory, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Haodong","family":"Duan","sequence":"additional","affiliation":[{"name":"Shanghai AI Laboratory, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Haian","family":"Huang","sequence":"additional","affiliation":[{"name":"Shanghai AI Laboratory, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yining","family":"Li","sequence":"additional","affiliation":[{"name":"Shanghai AI Laboratory, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kai","family":"Chen","sequence":"additional","affiliation":[{"name":"Shanghai AI Laboratory, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0417-234X","authenticated-orcid":false,"given":"Hua","family":"Yang","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Phi-3 technical report: A highly capable language model locally on your phone","volume-title":"arXiv:2404.14219","author":"Abdin","year":"2024"},{"key":"ref2","article-title":"Yi: Open foundation models by 01.AI","author":"Young","year":"2024","journal-title":"arXiv:2403.04652"},{"key":"ref3","volume-title":"Llama 3 Model Card","year":"2024"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.279"},{"key":"ref6","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":"ref7","article-title":"Language models are few-shot learners","author":"Brown","year":"2020","journal-title":"arXiv:2005.14165"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02506"},{"key":"ref9","article-title":"ALLaVA: Harnessing GPT4V-synthesized data for lite vision-language models","author":"Chen","year":"2024","journal-title":"arXiv:2402.11684"},{"key":"ref10","article-title":"Shikra: Unleashing multimodal LLM\u2019s referential dialogue magic","author":"Chen","year":"2023","journal-title":"arXiv:2306.15195"},{"key":"ref11","article-title":"ShareGPT4V: Improving large multi-modal models with better captions","author":"Chen","year":"2023","journal-title":"arXiv:2311.12793"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0850"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2023.3286896"},{"key":"ref14","article-title":"InternVL: Scaling up vision foundation models and aligning for generic visual-linguistic tasks","author":"Chen","year":"2023","journal-title":"arXiv:2312.14238"},{"key":"ref15","volume-title":"Vicuna: An Open-Source Chatbot Impressing GPT-4 With 90%* ChatGPT Quality","author":"Chiang","year":"2023"},{"issue":"240","key":"ref16","first-page":"1","article-title":"PaLM: Scaling language modeling with pathways","volume":"24","author":"Chowdhery","year":"2022","journal-title":"J. Mach. Learn. Res."},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73247-8_13"},{"issue":"70","key":"ref18","first-page":"1","article-title":"Scaling instruction-finetuned language models","volume":"25","author":"Chung","year":"2022","journal-title":"J. Mach. Learn. Res."},{"key":"ref19","volume-title":"XTuner: A Toolkit for Efficiently Fine-Tuning LLM","year":"2023"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2142"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/IVS.2019.8813788"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00544"},{"key":"ref23","article-title":"InternLM-XComposer2-4KHD: A pioneering large vision-language model handling resolutions from 336 pixels to 4K HD","author":"Dong","year":"2024","journal-title":"arXiv:2404.06512"},{"key":"ref24","first-page":"42566","article-title":"InternLM-XComposer2-4KHD: A pioneering large vision-language model handling resolutions from 336 pixels to 4K HD","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Dong"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1145\/3664647.3685520"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00720"},{"key":"ref27","article-title":"MiniCPM: Unveiling the potential of small language models with scalable training strategies","author":"Hu","year":"2024","journal-title":"arXiv:2404.06395"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/3dv66043.2025.00112"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00592"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46493-0_15"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00656"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00371"},{"key":"ref34","first-page":"71683","article-title":"OBELICS: An open web-scale filtered dataset of interleaved image-text documents","volume-title":"Proc. NeurIPS","volume":"36","author":"Lauren\u00e7on"},{"key":"ref35","article-title":"SEED-bench: Benchmarking multimodal LLMs with generative comprehension","author":"Li","year":"2023","journal-title":"arXiv:2307.16125"},{"key":"ref36","first-page":"19730","article-title":"BLIP-2: Bootstrapping language-image pre-training with frozen image encoders and large language models","volume-title":"Proc. ICML","author":"Li"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1007\/s11432-024-4321-9"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19812-0_42"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3453916"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02640"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00719"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.20"},{"key":"ref43","article-title":"Mini-gemini: Mining the potential of multi-modality vision language models","author":"Li","year":"2024","journal-title":"arXiv:2403.18814"},{"key":"ref44","article-title":"Monkey: Image resolution and text label are important things for large multi-modal models","author":"Li","year":"2023","journal-title":"arXiv:2311.06607"},{"key":"ref45","article-title":"Video-LLaVA: Learning united visual representation by alignment before projection","author":"Lin","year":"2023","journal-title":"arXiv:2311.10122"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref47","article-title":"SPHINX: The joint mixing of weights, tasks, and visual embeddings for multi-modal large language models","author":"Lin","year":"2023","journal-title":"arXiv:2311.07575"},{"key":"ref48","article-title":"HallusionBench: An advanced diagnostic suite for entangled language hallucination and visual illusion in large vision-language models","author":"Guan","year":"2023","journal-title":"arXiv:2310.14566"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2022.3165321"},{"key":"ref50","article-title":"Improved baselines with visual instruction tuning","author":"Liu","year":"2023","journal-title":"arXiv:2310.03744"},{"key":"ref51","article-title":"LLaVA-NeXT: Improved reasoning, OCR, and world knowledge","author":"Liu","year":"2024"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2021.3056725"},{"key":"ref54","article-title":"Grounding DINO: Marrying DINO with grounded pre-training for open-set object detection","author":"Liu","year":"2023","journal-title":"arXiv:2303.05499"},{"key":"ref55","article-title":"MMBench: Is your multi-modal model an all-around player?","author":"Liu","year":"2023","journal-title":"arXiv:2307.06281"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/TGRS.2020.2988782"},{"key":"ref57","article-title":"DeepSeek-VL: Towards real-world vision-language understanding","author":"Lu","year":"2024","journal-title":"arXiv:2403.05525"},{"key":"ref58","first-page":"2507","article-title":"Learn to explain: Multimodal reasoning via thought chains for science question answering","volume-title":"Proc. NeurIPS","volume":"35","author":"Lu"},{"key":"ref59","article-title":"Feast your eyes: Mixture-of-resolution adaptation for multimodal large language models","author":"Luo","year":"2024","journal-title":"arXiv:2403.03003"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1145\/3796716"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.679"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/WACV48630.2021.00225"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58598-3_24"},{"key":"ref64","first-page":"72983","article-title":"Scaling open-vocabulary object detection","volume-title":"Proc. NeurIPS","volume":"36","author":"Minderer"},{"key":"ref65","article-title":"GPT-4 technical report","volume-title":"arXiv:2303.08774","author":"Achiam","year":"2023"},{"key":"ref66","first-page":"27730","article-title":"Training language models to follow instructions with human feedback","volume-title":"Proc. NeurIPS","volume":"35","author":"Ouyang"},{"key":"ref67","article-title":"CogCoM: A visual language model with chain-of-manipulations reasoning","author":"Qi","year":"2024","journal-title":"arXiv:2402.04236"},{"key":"ref68","article-title":"Generalizable entity grounding via assistance of large language model","author":"Qi","year":"2024","journal-title":"arXiv:2402.02555"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00789"},{"key":"ref70","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"Proc. ICML","author":"Radford"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00690"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/tpami.2016.2577031"},{"key":"ref73","first-page":"25278","article-title":"LAION-5B: An open large-scale dataset for training next generation image-text models","volume-title":"Proc. NeurIPS","volume":"35","author":"Schuhmann"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P18-1238"},{"key":"ref75","article-title":"Eagle: Exploring the design space for multimodal LLMs with mixture of encoders","author":"Shi","year":"2024","journal-title":"arXiv:2408.15998"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00851"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2023.3234340"},{"key":"ref78","article-title":"Gemini: A family of highly capable multimodal models","author":"Team","year":"2023","journal-title":"arXiv:2312.11805"},{"key":"ref79","article-title":"InternLM: A multilingual language model with progressively enhanced capabilities","year":"2023"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-86331-9_50"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00914"},{"key":"ref82","article-title":"LLaMA: Open and efficient foundation language models","author":"Touvron","year":"2023","journal-title":"arXiv:2302.13971"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00956"},{"key":"ref84","article-title":"CogVLM: Visual expert for pretrained language models","author":"Wang","year":"2023","journal-title":"arXiv:2311.03079"},{"key":"ref85","article-title":"Lenna: Language enhanced reasoning detection assistant","author":"Wei","year":"2023","journal-title":"arXiv:2312.02433"},{"key":"ref86","article-title":"Towards semantic equivalence of tokenization in multimodal LLM","author":"Wu","year":"2024","journal-title":"arXiv:2406.05127"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.3233\/faia240541"},{"key":"ref88","article-title":"LLaVA-UHD: An LMM perceiving any aspect ratio and high-resolution images","author":"Xu","year":"2024","journal-title":"arXiv:2403.11703"},{"key":"ref89","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19836-6_31"},{"key":"ref90","first-page":"124","article-title":"Zero-shot video question answering via frozen bidirectional language models","volume-title":"Proc. NeurIPS","volume":"35","author":"Yang"},{"key":"ref91","article-title":"Baichuan 2: Open large-scale language models","author":"Yang","year":"2023","journal-title":"arXiv:2309.10305"},{"key":"ref92","article-title":"MPLUG-owl: Modularization empowers large language models with multimodality","author":"Ye","year":"2023","journal-title":"arXiv:2304.14178"},{"key":"ref93","article-title":"Ferret: Refer and ground anything anywhere at any granularity","author":"You","year":"2023","journal-title":"arXiv:2310.07704"},{"key":"ref94","article-title":"MM-Vet: Evaluating large multimodal models for integrated capabilities","author":"Yu","year":"2023","journal-title":"arXiv:2308.02490"},{"key":"ref95","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72775-7_24"},{"key":"ref96","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2023.3318672"},{"key":"ref97","article-title":"LLaVA-grounding: Grounded visual chat with large multimodal models","author":"Zhang","year":"2023","journal-title":"arXiv:2312.02949"},{"key":"ref98","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-demo.49"},{"key":"ref99","article-title":"InternLM-XComposer: A vision-language large model for advanced text-image comprehension and composition","author":"Zhang","year":"2023","journal-title":"arXiv:2309.15112"},{"key":"ref100","article-title":"LLaMA-adapter: Efficient fine-tuning of language models with zero-init attention","author":"Zhang","year":"2023","journal-title":"arXiv:2303.16199"},{"key":"ref101","article-title":"VL-uncertainty: Detecting hallucination in large vision-language model via uncertainty estimation","author":"Zhang","year":"2024","journal-title":"arXiv:2411.11919"},{"key":"ref102","article-title":"OPT: Open pre-trained transformer language models","author":"Zhang","year":"2022","journal-title":"arXiv:2205.01068"},{"key":"ref103","article-title":"OMG-LLaVA: Bridging image-level, object-level, pixel-level reasoning and understanding","author":"Zhang","year":"2024","journal-title":"arXiv:2406.19389"},{"key":"ref104","article-title":"Recognize anything: A strong image tagging model","author":"Zhang","year":"2023","journal-title":"arXiv:2306.03514"},{"key":"ref105","article-title":"Beyond LLaVA-HD: Diving into high-resolution large multimodal models","author":"Zhang","year":"2024","journal-title":"arXiv:2406.08487"},{"key":"ref106","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33019259"},{"key":"ref107","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00927"},{"key":"ref108","doi-asserted-by":"publisher","DOI":"10.1145\/3394171.3413896"},{"key":"ref109","article-title":"EdgeSAM: Prompt-in-the-loop distillation for SAM","author":"Zhou","year":"2023","journal-title":"arXiv:2312.06660"},{"key":"ref110","article-title":"MiniGPT-4: Enhancing vision-language understanding with advanced large language models","author":"Zhu","year":"2023","journal-title":"arXiv:2304.10592"},{"key":"ref111","article-title":"MoVA: Adapting mixture of vision experts to multimodal context","author":"Zong","year":"2024","journal-title":"arXiv:2404.13046"}],"container-title":["IEEE Transactions on Circuits and Systems for Video Technology"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/76\/11475579\/11299051.pdf?arnumber=11299051","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,7]],"date-time":"2026-04-07T20:03:23Z","timestamp":1775592203000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11299051\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4]]},"references-count":111,"journal-issue":{"issue":"4"},"URL":"https:\/\/doi.org\/10.1109\/tcsvt.2025.3643469","relation":{},"ISSN":["1051-8215","1558-2205"],"issn-type":[{"value":"1051-8215","type":"print"},{"value":"1558-2205","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4]]}}}