{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T10:23:28Z","timestamp":1777890208127,"version":"3.51.4"},"reference-count":98,"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":["62441231,62293542"],"award-info":[{"award-number":["62441231,62293542"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100004731","name":"Natural Science Foundation of Zhejiang Province","doi-asserted-by":"publisher","award":["LD25F020001"],"award-info":[{"award-number":["LD25F020001"]}],"id":[{"id":"10.13039\/501100004731","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100012166","name":"National Key R&D Program of China","doi-asserted-by":"publisher","award":["2022ZD0116302"],"award-info":[{"award-number":["2022ZD0116302"]}],"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\/iccv51701.2025.01953","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"21014-21025","source":"Crossref","is-referenced-by-count":1,"title":["EVEv2: Improved Baselines for Encoder-Free Vision-Language Models"],"prefix":"10.1109","author":[{"given":"Haiwen","family":"Diao","sequence":"first","affiliation":[{"name":"DLUT"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiaotong","family":"Li","sequence":"additional","affiliation":[{"name":"PKU"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yufeng","family":"Cui","sequence":"additional","affiliation":[{"name":"BAAI"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yueze","family":"Wang","sequence":"additional","affiliation":[{"name":"BAAI"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Haoge","family":"Deng","sequence":"additional","affiliation":[{"name":"BUPT"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ting","family":"Pan","sequence":"additional","affiliation":[{"name":"UCAS"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wenxuan","family":"Wang","sequence":"additional","affiliation":[{"name":"CASIA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Huchuan","family":"Lu","sequence":"additional","affiliation":[{"name":"DLUT"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xinlong","family":"Wang","sequence":"additional","affiliation":[{"name":"BAAI"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref2","article-title":"The claude 3 model family: Opus, sonnet, haiku","year":"2024","journal-title":"Claude-3 Model Card"},{"key":"ref3","author":"Bai","year":"2023","journal-title":"Qwen technical report"},{"issue":"2","key":"ref4","article-title":"Qwen-vl: A versatile vision-language model for understanding, localization, text reading, and beyond","volume":"1","author":"Bai","year":"2023","journal-title":"arXiv preprint"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2384"},{"key":"ref6","author":"Bavishi","year":"2023","journal-title":"Introducing our multimodal models"},{"key":"ref7","article-title":"Paligemma: A versatile 3b vlm for transfer","author":"Beyer","year":"2024","journal-title":"arXiv preprint"},{"key":"ref8","author":"Bi","year":"2024","journal-title":"Deepseek 11m: Scaling open-source language models with longtermism"},{"key":"ref9","author":"Cai","year":"2024","journal-title":"Internlm2 technical report"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00951"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72643-9_22"},{"key":"ref12","article-title":"A single transformer for scalable vision-language modeling","author":"Chen","year":"2024","journal-title":"arXiv preprint"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.02283"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1007\/s11432-024-4231-5"},{"key":"ref15","article-title":"Xing","author":"Chiang","year":"2023","journal-title":"Vicuna: An open-source chatbot impressing gpt-4 with 90 \\%^{*} chatgpt quality"},{"key":"ref16","first-page":"7480","article-title":"Scaling vision transformers to 22 billion parameters","author":"Dehghani","year":"2023","journal-title":"ICML"},{"key":"ref17","article-title":"Unveiling encoder-free vision-language models","author":"Diao","year":"2024","journal-title":"arXiv preprint"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02714"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72784-9_5"},{"key":"ref20","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","author":"Dosovitskiy","year":"2021","journal-title":"ICLR"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01268"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01855"},{"key":"ref23","author":"Fu","year":"2023","journal-title":"MME: A comprehensive evaluation benchmark for multimodal large language models"},{"key":"ref24","article-title":"Datacomp: In search of the next generation of multimodal datasets","author":"Yitzhak Gadre","year":"2024","journal-title":"NeurIPS, 36"},{"key":"ref25","author":"Gao","year":"2023","journal-title":"Llama-adapter v 2: Parameter-efficient visual instruction model"},{"key":"ref26","article-title":"Making LLaMA SEE and draw with SEED tokenizer","volume-title":"The Twelfth International Conference on Learning Representations","author":"Ge","year":"2024"},{"key":"ref27","article-title":"Infinity-mm: Scaling multimodal performance with large-scale and high-quality instruction data","author":"Gu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.01354"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"ref30","article-title":"IDEFICS Research Team","volume-title":"Introducing idefics: An open reproduction of state-of-the-art visual language model","year":"2023"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46493-0_15"},{"key":"ref32","article-title":"Adam: A method for stochastic optimization","author":"Kingma","year":"2015","journal-title":"ICLR"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/iccv51070.2023.00371"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-020-01316-z"},{"key":"ref35","author":"Li","year":"2023","journal-title":"Seed-bench: Benchmarking multimodal 11 ms with generative comprehension"},{"key":"ref36","article-title":"Llava-onevision: Easy visual task transfer","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref37","article-title":"Aria: An open multimodal native mixture-of-experts model","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref38","first-page":"12888","article-title":"H. Hoi. BLIP: bootstrapping language-image pre-training for unified vision-language understanding and generation","author":"Li","year":"2022","journal-title":"ICLR"},{"key":"ref39","first-page":"19730","article-title":"BLIP-2 : bootstrapping language-image pre-training with frozen image encoders and large language models","author":"Li","year":"2023","journal-title":"ICML"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20056-4_14"},{"key":"ref41","article-title":"Imagefolder: Autoregressive image generation with folded tokens","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref42","article-title":"Densefusion-1m: Merging vision experts for comprehensive multimodal perception","author":"Li","year":"2024","journal-title":"arXiv preprint"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.20"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.02527"},{"key":"ref45","article-title":"Moma: Efficient early-fusion pre-training with mixture of modality-aware experts","author":"Victoria Lin","year":"2024","journal-title":"arXiv preprint"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.02484"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref48","author":"Liu","year":"2024","journal-title":"Llava-next: Improved reasoning, ocr, and world knowledge"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1145\/3539597.3573023"},{"key":"ref50","article-title":"On the hidden mystery of ocr in large multimodal models","author":"Liu","year":"2023","journal-title":"arXiv preprint"},{"key":"ref51","author":"Lu","year":"2024","journal-title":"Deepseek-vl: Towards real-world visionlanguage understanding"},{"key":"ref52","article-title":"Learn to explain: Multimodal reasoning via thought chains for science question answering","author":"Lu","year":"2022","journal-title":"NeurIPS"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.02324"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.findings-acl.177"},{"key":"ref55","author":"Report","journal-title":"\u00a0 2023"},{"key":"ref56","author":"Qu","year":"2024","journal-title":"Tokenflow: Unified image tokenizer for multimodal understanding and generation. arXiv preprint arXiv:2412.03069"},{"key":"ref57","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","author":"Radford","year":"2021","journal-title":"ICML"},{"key":"ref58","first-page":"25278","article-title":"Laion-5b: An open large-scale dataset for training next generation image-text models","volume":"35","author":"Schuhmann","year":"2022","journal-title":"NeurIPS"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00851"},{"key":"ref60","article-title":"Autoregressive model beats diffusion: Llama for scalable image generation","author":"Sun","year":"2024","journal-title":"arXiv preprint"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.01365"},{"key":"ref62","author":"Sun","year":"2023","journal-title":"EVA-CLIP: improved training techniques for CLIP at scale"},{"key":"ref63","author":"Sun","year":"2023","journal-title":"Generative pretraining in multimodality"},{"key":"ref64","author":"Sun","year":"2024","journal-title":"EVA-CLI P-18B: scaling CLIP to 18 billion parameters"},{"key":"ref65","article-title":"Chameleon: Mixed-modal early-fusion foundation models","author":"Team","year":"2024","journal-title":"arXiv preprint"},{"key":"ref66","author":"Team","year":"2023","journal-title":"Gemini: a family of highly capable multimodal models"},{"key":"ref67","article-title":"InternLM Team","volume-title":"Internlm: A multilingual language model with progressively enhanced capabilities","year":"2023"},{"key":"ref68","year":"2024","journal-title":"Llama 3.2: Revolutionizing edge ai and vision with open, customizable models"},{"key":"ref69","year":"2024","journal-title":"Qwen2.5: A party of foundation models"},{"key":"ref70","article-title":"Cambrian-1: A fully open, vision-centric exploration of multimodal 11 ms","author":"Tong","year":"2024","journal-title":"arXiv preprint"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00914"},{"key":"ref72","author":"Touvron","year":"2023","journal-title":"Llama: Open and efficient foundation language models"},{"key":"ref73","author":"Touvron","year":"2023","journal-title":"Llama 2: Open foundation and fine-tuned chat models"},{"key":"ref74","article-title":"Neural discrete representation learning","author":"Van","year":"2017","journal-title":"NeurIPS"},{"key":"ref75","article-title":"Attention is all you need","author":"Vaswani","year":"2017","journal-title":"NIPS"},{"key":"ref76","article-title":"Qwen2-vl: Enhancing vision-language model\u2019s perception of the world at any resolution","author":"Wang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52729.2023.01838"},{"key":"ref78","article-title":"Emu 3: Next-token prediction is all you need","author":"Wang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.emnlp-main.255"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.01210"},{"key":"ref81","article-title":"Vila-u: a unified foundation model integrating visual understanding and generation","author":"Wu","year":"2024","journal-title":"arXiv preprint"},{"key":"ref82","article-title":"x.ai","year":"2024","journal-title":"Grok-1.5 vision preview"},{"key":"ref83","article-title":"Show-o: One single transformer to unify multimodal understanding and generation","author":"Xie","year":"2024","journal-title":"arXiv preprint"},{"key":"ref84","article-title":"Musevl: Modeling unified vlm through semantic discrete encoding","author":"Xie","year":"2024","journal-title":"arXiv preprint"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73010-8_23"},{"key":"ref86","article-title":"xgen-mm (blip-3): A family of open large multimodal models","author":"Xue","year":"2024","journal-title":"arXiv preprint"},{"key":"ref87","article-title":"Qwen2 technical report","author":"Yang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref88","author":"Yang","year":"2023","journal-title":"The dawn of lmms: Preliminary explorations with gpt-4v (ision)"},{"key":"ref89","author":"Ye","year":"2023","journal-title":"mplug-owl: Modularization empowers large language models with multimodality"},{"key":"ref90","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.01239"},{"key":"ref91","author":"Yu","year":"2023","journal-title":"Mm -vet: Evaluating large multimodal models for integrated capabilities"},{"key":"ref92","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.00913"},{"key":"ref93","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01100"},{"key":"ref94","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.521"},{"key":"ref95","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.findings-naacl.51"},{"key":"ref96","article-title":"Llamaadapter: Efficient fine-tuning of language models with zeroinit attention","author":"Zhang","year":"2024","journal-title":"ICLR"},{"key":"ref97","first-page":"23412","article-title":"Movq: Modulating quantized vectors for high-fidelity image generation","volume":"35","author":"Zheng","year":"2022","journal-title":"NeurIPS"},{"key":"ref98","author":"Zheng","year":"2023","journal-title":"Efficiently programming large language models using sglang. arXiv: 2312.07104"}],"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\/11445103.pdf?arnumber=11445103","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T05:31:35Z","timestamp":1777613495000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11445103\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":98,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.01953","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}