{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T10:21:19Z","timestamp":1777890079433,"version":"3.51.4"},"reference-count":91,"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\/iccv51701.2025.01554","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"16733-16744","source":"Crossref","is-referenced-by-count":0,"title":["X2i: Seamless Integration of Multimodal Understanding Into Diffusion Transformer Via Attention Distillation"],"prefix":"10.1109","author":[{"given":"Jian","family":"Ma","sequence":"first","affiliation":[{"name":"OPPO AI Center"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Qirong","family":"Peng","sequence":"additional","affiliation":[{"name":"OPPO AI Center"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xu","family":"Guo","sequence":"additional","affiliation":[{"name":"Tsinghua University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Chen","family":"Chen","sequence":"additional","affiliation":[{"name":"OPPO AI Center"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Haonan","family":"Lu","sequence":"additional","affiliation":[{"name":"OPPO AI Center"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhenyu","family":"Yang","sequence":"additional","affiliation":[{"name":"OPPO AI Center"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","volume-title":"Phi-3 technical report: A highly capable language model locally on your phone","author":"Abdin","year":"2024"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1723"},{"key":"ref3","first-page":"1","author":"Bai","year":"2023","journal-title":"Qwen-vl: A versatile vision-language model for understanding, localization, text reading, and beyond"},{"key":"ref4","volume-title":"Qwen2.5-vl technical report","author":"Bai","year":"2025"},{"key":"ref5","author":"Balaji","year":"2023","journal-title":"ediff-i: Text-to-image diffusion models with an ensemble of expert denoisers"},{"key":"ref6","author":"Berman","year":"2024","journal-title":"Mumu: Bootstrapping multimodal image generation from text-toimage data"},{"key":"ref7","author":"Chen","year":"2023","journal-title":"X-llm: Bootstrapping advanced large language models by treating multi-modalities as foreign languages"},{"key":"ref8","first-page":"1","author":"Chen","year":"2023","journal-title":"Pixart-a: Fast training of diffusion transformer for photorealistic text-to-image synthesis"},{"key":"ref9","article-title":"Cohen","author":"Chen","year":"2023","journal-title":"Subject-driven text-to-image generation via apprenticeship learning"},{"key":"ref10","first-page":"1","article-title":"Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks","author":"Chen","year":"2023","journal-title":"arXiv preprint"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52729.2023.00276"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72848-8_1"},{"key":"ref13","year":"2024","journal-title":"Deepseek-v2: A strong, economical, and efficient mixture-of-experts language model"},{"key":"ref14","author":"Ding","year":"2025","journal-title":"Efficient-vdit: Efficient video diffusion transformers with attention tile"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.00746"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01268"},{"key":"ref17","first-page":"1","author":"Esser","year":"2024","journal-title":"Scaling rectified flow transformers for high-resolution image synthesis"},{"key":"ref18","author":"Fu","year":"2024","journal-title":"Guiding instruction-based image editing via multimodal large language models"},{"key":"ref19","author":"Gao","year":"2024","journal-title":"Lumina-t2x: Transforming text into any modality, resolution, and duration via flow-based large diffusion transformers"},{"key":"ref20","author":"Gu","year":"2024","journal-title":"Minillm: Knowledge distillation of large language models"},{"key":"ref21","first-page":"5","author":"Han","year":"2024","journal-title":"Evalmuse-40k: A reliable and fine-grained benchmark with comprehensive human annotations for text-to-image generation model evaluation"},{"key":"ref22","first-page":"1","volume-title":"Ace: Allround creator and editor following instructions via diffusion transformer","author":"Han","year":"2024"},{"key":"ref23","author":"Hu","year":"2021","journal-title":"Lora: Low-rank adaptation of large language models"},{"key":"ref24","author":"Hu","year":"2024","journal-title":"Minicpm: Unveiling the potential of small language models with scalable training strategies"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2025.3531907"},{"key":"ref26","author":"Jiang","year":"2024","journal-title":"Diffartist: Towards aesthetic-aligned diffusion model control for training-free text-driven stylization"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00453"},{"key":"ref28","author":"Kim","year":"2024","journal-title":"Consistency trajectory models: Learning probability flow ode trajectory of diffusion"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1594"},{"key":"ref30","author":"Kong","year":"2025","journal-title":"Hunyuanvideo: A systematic framework for large video generative models"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00192"},{"key":"ref32","first-page":"1","volume-title":"Labs. Flux","author":"Forest","year":"2024"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/cvprw63382.2024.00538"},{"key":"ref34","first-page":"1","article-title":"Blip-diffusion: Pretrained subject representation for controllable text-to-image generation and editing","volume":"36","author":"Li","year":"2024","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref35","first-page":"1","author":"Li","year":"2023","journal-title":"Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models"},{"key":"ref36","author":"Li","year":"2024","journal-title":"Hunyuan-dit: A powerful multi-resolution diffusion transformer with fine-grained chinese understanding"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72673-6_20"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1145\/3590140.3592851"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73226-3_21"},{"key":"ref42","author":"Lu","year":"2024","journal-title":"Deepseek-vl: Towards real-world visionlanguage understanding"},{"key":"ref43","first-page":"2","author":"Lu","year":"2024","journal-title":"Qwen2vl-flux: Unifying image and text guidance for controllable image generation"},{"key":"ref44","author":"Ma","year":"2024","journal-title":"Exploring the role of large language models in prompt encoding for diffusion models"},{"key":"ref45","first-page":"1","article-title":"Peadiffusion: Parameter-efficient adapter with knowledge distillation in non-english text-to-image generation","author":"Ma","year":"2023","journal-title":"arXiv preprint"},{"key":"ref46","article-title":"Glyphdraw: Learning to draw chinese characters in image synthesis models coherently","author":"Ma","year":"2023","journal-title":"arXiv preprint"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1145\/3641519.3657469"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i6.32636"},{"key":"ref49","author":"Mou","year":"2023","journal-title":"T2iadapter: Learning adapters to dig out more controllable ability for text-to-image diffusion models"},{"key":"ref50","article-title":"Glide: Towards photorealistic image generation and editing with text-guided diffusion models","author":"Nichol","year":"2021","journal-title":"arXiv preprint"},{"key":"ref51","first-page":"3","author":"Pan","year":"2024","journal-title":"Kosmos-g: Generating images in context with multimodal large language models"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00387"},{"key":"ref53","author":"Peng","year":"2024","journal-title":"Controlnext: Powerful and efficient control for image and video generation"},{"key":"ref54","first-page":"1","author":"Podell","year":"2023","journal-title":"Sdxl: Improving latent diffusion models for high-resolution image synthesis"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.02110"},{"key":"ref56","author":"Qin","year":"2023","journal-title":"Unicontrol: A unified diffusion model for controllable visual generation in the wild"},{"key":"ref57","author":"Radford","year":"2021","journal-title":"Learning transferable visual models from natural language supervision"},{"key":"ref58","article-title":"Liu","author":"Raffel","year":"2023","journal-title":"Exploring the limits of transfer learning with a unified text-to-text transformer"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/SC41405.2020.00024"},{"key":"ref60","author":"Ramesh","year":"2021","journal-title":"Zero-shot text-to-image generation"},{"issue":"2","key":"ref61","first-page":"3","article-title":"Hierarchical text-conditional image generation with clip latents","volume":"1","author":"Ramesh","year":"2022","journal-title":"arXiv preprint"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01042"},{"key":"ref63","first-page":"234","article-title":"Unet: Convolutional networks for biomedical image segmentation","volume-title":"In Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015: 18th International Conference, Munich, Germany, October 5\u20139, 2015, Proceedings, Part III 18","author":"Ronneberger"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02155"},{"key":"ref65","first-page":"3","author":"Saharia","year":"2022","journal-title":"Photorealistic text-to-image diffusion models with deep language understanding"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2643"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1833"},{"key":"ref68","first-page":"1","author":"Shi","year":"2024","journal-title":"Seededit: Align image re-generation to image editing"},{"key":"ref69","author":"Skean","year":"2024","journal-title":"Does representation matter? exploring intermediate layers in large language models"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51701.2025.01386"},{"key":"ref71","first-page":"1","article-title":"Kolors: Effective training of diffusion model for photorealistic text-to-image synthesis","year":"2024","journal-title":"arXiv preprint"},{"key":"ref72","author":"Touvron","year":"2023","journal-title":"Llama 2: Open foundation and fine-tuned chat models"},{"key":"ref73","author":"Wang","year":"2022","journal-title":"Attention distillation: self-supervised vision transformer students need more guidance"},{"key":"ref74","author":"Wang","year":"2024","journal-title":"Qwen2-vl: Enhancing vision-language model\u2019s perception of the world at any resolution"},{"key":"ref75","author":"Wang","year":"2024","journal-title":"Cogvlm: Visual expert for pretrained language models"},{"key":"ref76","author":"Wu","year":"2024","journal-title":"Next-gpt: Any-to-any multimodal 11 m"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/iccv51070.2023.00200"},{"key":"ref78","author":"Wu","year":"2024","journal-title":"Deepseek-vl2: Mixture-of-experts vision-language models for advanced multimodal understanding"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.02657"},{"key":"ref80","author":"Xie","year":"2024","journal-title":"Sana: Efficient high-resolution image synthesis with linear diffusion transformers"},{"key":"ref81","author":"Xu","year":"2023","journal-title":"Imagereward: Learning and evaluating human preferences for text-to-image generation"},{"key":"ref82","author":"Xue","year":"2024","journal-title":"Raphael: Text-to-image generation via large mixture of diffusion paths"},{"key":"ref83","author":"Yang","year":"2023","journal-title":"Controllable textual inversion for personalized text-to-image generation"},{"key":"ref84","first-page":"1","article-title":"Ipadapter: Text compatible image prompt adapter for text-toimage diffusion models","author":"Ye","year":"2023","journal-title":"arXiv preprint"},{"key":"ref85","author":"Yu","year":"2022","journal-title":"Scaling autoregressive models for content-rich text-to-image generation"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.02433"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.521"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00355"},{"key":"ref89","author":"Zheng","year":"2024","journal-title":"Trajectory consistency distillation: Improved latent consistency distillation by semi-linear consistency function with trajectory mapping"},{"key":"ref90","author":"Zhu","year":"2024","journal-title":"Slimflow: Training smaller one-step diffusion models with rectified flow"},{"key":"ref91","author":"Zong","year":"2024","journal-title":"Easyref: Omni-generalized group image reference for diffusion models via multimodal 1lm"}],"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\/11444032.pdf?arnumber=11444032","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T05:28:50Z","timestamp":1777613330000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11444032\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":91,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.01554","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}