{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,5,22]],"date-time":"2025-05-22T06:10:13Z","timestamp":1747894213696,"version":"3.41.0"},"reference-count":28,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,3,31]],"date-time":"2025-03-31T00:00:00Z","timestamp":1743379200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,3,31]],"date-time":"2025-03-31T00:00:00Z","timestamp":1743379200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100012166","name":"National Key R&D Program of China","doi-asserted-by":"publisher","award":["2022ZD0160304"],"award-info":[{"award-number":["2022ZD0160304"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100009592","name":"Beijing Municipal Science and Technology Project","doi-asserted-by":"publisher","award":["Z231100010323002"],"award-info":[{"award-number":["Z231100010323002"]}],"id":[{"id":"10.13039\/501100009592","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,3,31]]},"DOI":"10.23919\/date64628.2025.10993131","type":"proceedings-article","created":{"date-parts":[[2025,5,21]],"date-time":"2025-05-21T17:36:35Z","timestamp":1747848995000},"page":"1-7","source":"Crossref","is-referenced-by-count":0,"title":["BMP-SD: Marrying Binary and Mixed-Precision Quantization for Efficient Stable Diffusion Inference"],"prefix":"10.23919","author":[{"given":"Cheng","family":"Gu","sequence":"first","affiliation":[{"name":"Shanghai Jiao Tong University,Department of Computer Science and Engineering,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Gang","family":"Li","sequence":"additional","affiliation":[{"name":"CASIA,The Key Laboratory of Cognition and Decision Intelligence for Complex Systems,Beijing,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiaolong","family":"Lin","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University,Department of Computer Science and Engineering,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jiayao","family":"Ling","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University,Department of Computer Science and Engineering,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jian","family":"Cheng","sequence":"additional","affiliation":[{"name":"CASIA,The Key Laboratory of Cognition and Decision Intelligence for Complex Systems,Beijing,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiaoyao","family":"Liang","sequence":"additional","affiliation":[{"name":"Shanghai Jiao Tong University,Department of Computer Science and Engineering,Shanghai,China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01042"},{"key":"ref2","article-title":"Generative adversarial nets","volume":"27","author":"Goodfellow","year":"2014","journal-title":"Advances in neural information processing systems"},{"key":"ref3","volume-title":"Advances in neural information processing systems 17: proceedings of the 2004 conference","volume":"17","author":"Saul","year":"2005"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2022.3204461"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00582"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01117"},{"key":"ref7","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":"ref8","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"ref9","article-title":"Auto-encoding variational bayes","author":"Kingma","year":"2013","journal-title":"arXiv preprint"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01608"},{"key":"ref11","article-title":"Quest: Low-bit diffusion model quantization via efficient selective finetuning","author":"Wang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3664647.3681451"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/DAC56929.2023.10247662"},{"key":"ref15","doi-asserted-by":"crossref","first-page":"304","DOI":"10.1145\/3307650.3322255","article-title":"Laconic deep learning inference acceleration","volume-title":"Proceedings of the 46th International Symposium on Computer Architecture","author":"Sharify","year":"2019"},{"key":"ref16","doi-asserted-by":"crossref","first-page":"382","DOI":"10.1145\/3123939.3123982","article-title":"Bit-pragmatic deep neural network computing","volume-title":"Proceedings of the 50th annual IEEE\/ACM international symposium on microarchitecture","author":"Albericio","year":"2017"},{"volume-title":"Markov chains","year":"1998","author":"Norris","key":"ref17"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3261988"},{"key":"ref19","article-title":"Layer normalization","author":"Lei Ba","year":"2016","journal-title":"ArXiv e-prints"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01261-8_1"},{"key":"ref21","article-title":"Attention is all you need","author":"Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"journal-title":"arXiv preprint","article-title":"Glu variants improve transformer","author":"Shazeer","key":"ref22"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2017.12.012"},{"key":"ref24","article-title":"Billm: Pushing the limit of post-training quantization for llms","author":"Huang","year":"2024","journal-title":"arXiv preprint"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1145\/3234150"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72630-9_17"},{"key":"ref27","article-title":"Bitsfusion: 1.99 bits weight quantization of diffusion model","author":"Sui","year":"2024","journal-title":"arXiv preprint"},{"key":"ref28","article-title":"Omniquant: Omnidirectionally calibrated quantization for large language models","author":"Shao","year":"2023","journal-title":"arXiv preprint"}],"event":{"name":"2025 Design, Automation & Test in Europe Conference (DATE)","start":{"date-parts":[[2025,3,31]]},"location":"Lyon, France","end":{"date-parts":[[2025,4,2]]}},"container-title":["2025 Design, Automation & Test in Europe Conference (DATE)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/10992638\/10992588\/10993131.pdf?arnumber=10993131","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,22]],"date-time":"2025-05-22T05:33:11Z","timestamp":1747891991000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10993131\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,31]]},"references-count":28,"URL":"https:\/\/doi.org\/10.23919\/date64628.2025.10993131","relation":{},"subject":[],"published":{"date-parts":[[2025,3,31]]}}}