{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,17]],"date-time":"2026-07-17T06:06:37Z","timestamp":1784268397016,"version":"3.55.0"},"reference-count":121,"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\/iccv51701.2025.01414","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"15240-15252","source":"Crossref","is-referenced-by-count":2,"title":["Adaptive Caching for Faster Video Generation With Diffusion Transformers"],"prefix":"10.1109","author":[{"given":"Kumara","family":"Kahatapitiya","sequence":"first","affiliation":[{"name":"Meta AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Haozhe","family":"Liu","sequence":"additional","affiliation":[{"name":"Meta AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sen","family":"He","sequence":"additional","affiliation":[{"name":"Meta AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ding","family":"Liu","sequence":"additional","affiliation":[{"name":"Meta AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Menglin","family":"Jia","sequence":"additional","affiliation":[{"name":"Meta AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chenyang","family":"Zhang","sequence":"additional","affiliation":[{"name":"Meta AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael S.","family":"Ryoo","sequence":"additional","affiliation":[{"name":"Stony Brook University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tian","family":"Xie","sequence":"additional","affiliation":[{"name":"Meta AI"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Adir: Adaptive diffusion for image reconstruction","author":"Abu-Hussein","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.02637"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01767"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3592450"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19784-0_41"},{"key":"ref6","article-title":"Stable video diffusion: Scaling latent video diffusion models to large datasets","author":"Blattmann","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02161"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW59228.2023.00484"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01103"},{"key":"ref10","article-title":"Muse: Text-to-image generation via masked generative transformers","author":"Chang","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00698"},{"key":"ref12","article-title":"Pixart-alpha: Fast training of diffusion transformer for photorealistic text-to-image synthesis","author":"Chen","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73411-3_5"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.02636"},{"key":"ref15","article-title":"Delta-dit: A training-free acceleration method tailored for diffusion transformers","author":"Chen","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref16","article-title":"Follow-your-canvas: Higher-resolution video outpainting with extensive content generation","author":"Chen","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref17","article-title":"Seine: Short-to-long video diffusion model for generative transition and prediction","volume-title":"The Twelfth International Conference on Learning Representations","author":"Chen","year":"2023"},{"key":"ref18","article-title":"Adversarial video generation on complex datasets","author":"Clark","year":"2019","journal-title":"arXiv preprint arXiv"},{"key":"ref19","article-title":"Diffedit: Diffusion-based semantic image editing with mask guidance","author":"Couairon","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref20","article-title":"Emu: Enhancing image generation models using photogenic needles in a haystack","author":"Dai","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i15.33782"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3581783.3612478"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73229-4_22"},{"key":"ref24","article-title":"Lumina-t2x: Transforming text into any modality, resolution, and duration via flow-based large diffusion transformers","author":"Gao","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref25","article-title":"Emu video: Factorizing text-to-video generation by explicit image conditioning","author":"Girdhar","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/3422622"},{"key":"ref27","article-title":"I2v-adapter: A general image-to-video adapter for video diffusion models","author":"Guo","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00798"},{"key":"ref29","article-title":"Latent video diffusion models for high-fidelity long video generation","author":"He","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0580"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.00245"},{"key":"ref32","article-title":"Prompt-to-prompt image editing with cross attention control","author":"Hertz","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref33","first-page":"6840","article-title":"Denoising diffusion probabilistic models","volume":"33","author":"Ho","year":"2020","journal-title":"Advances in neural information processing systems"},{"key":"ref34","article-title":"Imagen video: High definition video generation with diffusion models","author":"Ho","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref35","article-title":"Cogvideo: Large-scale pretraining for text-to-video generation via transformers","author":"Hong","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.00193"},{"key":"ref37","first-page":"13916","article-title":"Make-an-audio: Text-to-audio generation with prompt-enhanced diffusion models","volume-title":"International Conference on Machine Learning","author":"Huang"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02060"},{"key":"ref39","article-title":"Objectcentric diffusion for efficient video editing","author":"Kahatapitiya","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00453"},{"key":"ref41","article-title":"Auto-encoding variational bayes","author":"P Kingma","year":"2013","journal-title":"arXiv preprint arXiv"},{"key":"ref42","article-title":"Kling AI","volume-title":"Next-generation ai creative studio","year":"2024"},{"key":"ref43","article-title":"Videopoet: A large language model for zero-shot video generation","author":"Kondratyuk","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref44","article-title":"Diffwave: A versatile diffusion model for audio synthesis","author":"Kong","year":"2020","journal-title":"arXiv preprint arXiv"},{"key":"ref45","volume-title":"Open-sora-plan","author":"Lab","year":"2024"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01608"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00715"},{"key":"ref48","article-title":"Snapfusion: Text-to-image diffusion model on mobile devices within two seconds","volume":"36","author":"Li","year":"2024","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref49","article-title":"Mardini: Masked autoregressive diffusion for video generation at scale","author":"Liu","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/cvprw67362.2025.00305"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-19790-1_26"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1109\/iccv51070.2023.00853"},{"key":"ref53","article-title":"Instaflow: One step is enough for high-quality diffusionbased text-to-image generation","volume-title":"The Twelfth International Conference on Learning Representations","author":"Liu","year":"2023"},{"key":"ref54","article-title":"Sora: A review on background, technology, limitations, and opportunities of large vision models","author":"Liu","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1007\/s11633-025-1562-4"},{"key":"ref56","article-title":"Vdt: General-purpose video diffusion transformers via mask modeling","author":"Lu","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref57","article-title":"Luma AI","volume-title":"Dream machine","year":"2024"},{"key":"ref58","article-title":"Latent consistency models: Synthesizing highresolution images with few-step inference","author":"Luo","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref59","article-title":"Fastercache: Training-free video diffusion model acceleration with high quality","author":"Lv","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72980-5_2"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01492"},{"key":"ref62","article-title":"Latte: Latent diffusion transformer for video generation","author":"Ma","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.01973"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01374"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72691-0_4"},{"key":"ref66","article-title":"OpenAI","volume-title":"Video generation models as world simulators","year":"2024"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00387"},{"key":"ref68","article-title":"Sdxl: Improving latent diffusion models for high-resolution image synthesis","author":"Podell","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref69","article-title":"Dreamfusion: Text-to-3d using 2d diffusion","author":"Poole","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01460"},{"issue":"2","key":"ref71","article-title":"Hierarchical text-conditional image generation with clip latents","volume":"1","author":"Ramesh","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref72","article-title":"Discrete variational autoencoders","author":"Tyler Rolfe","year":"2016","journal-title":"arXiv preprint arXiv"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01042"},{"key":"ref74","first-page":"234","article-title":"Unet: Convolutional networks for biomedical image segmentation","volume-title":"Medical image computing and computer-assisted intervention-MICCAI 2015: 18th international conference","author":"Ronneberger","year":"2015"},{"key":"ref75","article-title":"Runway AI","volume-title":"Gen-3 alpha: A new frontier for video generation","year":"2024"},{"key":"ref76","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2643"},{"key":"ref77","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.308"},{"key":"ref78","article-title":"Progressive distillation for fast sampling of diffusion models","author":"Salimans","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref79","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73016-0_6"},{"key":"ref80","article-title":"Fora: Fast-forward caching in diffusion transformer acceleration","author":"Selvaraju","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref81","article-title":"Make-a-video: Text-to-video generation without text-video data","author":"Singer","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref82","article-title":"Denoising diffusion implicit models","author":"Song","year":"2020","journal-title":"arXiv preprint arXiv"},{"key":"ref83","article-title":"Consistency models","author":"Song","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00886"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1109\/WACV57701.2024.00622"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2012.2221191"},{"key":"ref87","article-title":"Video-infinity: Distributed long video generation","author":"Tan","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref88","article-title":"Deediff: Dynamic uncertaintyaware early exiting for accelerating diffusion model generation","author":"Tang","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref89","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00165"},{"key":"ref90","article-title":"Phenaki: Variable length video generation from open domain textual descriptions","volume-title":"International Conference on Learning Representations","author":"Villegas","year":"2022"},{"key":"ref91","article-title":"Generating videos with scene dynamics","volume":"29","author":"Vondrick","year":"2016","journal-title":"Advances in neural information processing systems"},{"key":"ref92","article-title":"Gen-l-video: Multi-text to long video generation via temporal co-denoising","author":"Wang","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref93","article-title":"Be-your-outpainter: Mastering video outpainting through input-specific adaptation","author":"Wang","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref94","article-title":"Quest: Low-bit diffusion model quantization via efficient selective finetuning","author":"Wang","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref95","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01761"},{"key":"ref96","article-title":"Generative inbetweening: Adapting image-to-video models for keyframe interpolation","author":"Wang","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref97","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2003.815165"},{"key":"ref98","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.00594"},{"key":"ref99","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00701"},{"key":"ref100","article-title":"Customcrafter: Customized video generation with preserving motion and concept composition abilities","author":"Wu","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref101","article-title":"Show-o: One single transformer to unify multimodal understanding and generation","author":"Xie","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref102","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02148"},{"key":"ref103","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72952-2_23"},{"key":"ref104","article-title":"Camco: Camera-controllable 3d-consistent image-to-video generation","author":"Xu","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref105","article-title":"Videogpt: Video generation using vq-vae and transformers","author":"Yan","year":"2021","journal-title":"arXiv preprint arXiv"},{"key":"ref106","doi-asserted-by":"publisher","DOI":"10.1145\/3610548.3618160"},{"key":"ref107","article-title":"Compositional video generation as flow equalization","author":"Yang","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref108","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02160"},{"key":"ref109","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00809"},{"key":"ref110","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.acl-long.73"},{"key":"ref111","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00632"},{"key":"ref112","article-title":"Generating videos with dynamics-aware implicit generative adversarial networks","author":"Yu","year":"2022","journal-title":"arXiv preprint arXiv"},{"key":"ref113","article-title":"Crossattention makes inference cumbersome in text-to-image diffusion models","author":"Zhang","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref114","article-title":"Tora: Trajectory-oriented diffusion transformer for video generation","author":"Zhang","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref115","article-title":"Moviedreamer: Hierarchical generation for coherent long visual sequence","author":"Zhao","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref116","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72992-8_16"},{"key":"ref117","article-title":"Dsp: Dynamic sequence parallelism for multi-dimensional transformers","author":"Zhao","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref118","article-title":"Real-time video generation with pyramid attention broadcast","author":"Zhao","year":"2024","journal-title":"arXiv preprint arXiv"},{"key":"ref119","article-title":"Mobilediffusion: Subsecond text-to-image generation on mobile devices","author":"Zhao","year":"2023","journal-title":"arXiv preprint arXiv"},{"key":"ref120","volume-title":"Open-sora: Democratizing efficient video production for all","author":"Zheng","year":"2024"},{"key":"ref121","article-title":"Storydiffusion: Consistent selfattention for long-range image and video generation","author":"Zhou","year":"2024","journal-title":"arXiv preprint arXiv"}],"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\/11445977.pdf?arnumber=11445977","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T05:33:05Z","timestamp":1777613585000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11445977\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":121,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.01414","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}