{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T06:17:20Z","timestamp":1783405040466,"version":"3.54.6"},"reference-count":41,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"8","license":[{"start":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T00:00:00Z","timestamp":1785542400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T00:00:00Z","timestamp":1785542400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T00:00:00Z","timestamp":1785542400000},"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":["U2333211"],"award-info":[{"award-number":["U2333211"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. on Mobile Comput."],"published-print":{"date-parts":[[2026,8]]},"DOI":"10.1109\/tmc.2026.3667591","type":"journal-article","created":{"date-parts":[[2026,2,24]],"date-time":"2026-02-24T20:59:58Z","timestamp":1771966798000},"page":"11537-11553","source":"Crossref","is-referenced-by-count":0,"title":["EdgeSD: Efficient Speculative Decoding With Vision-Decoding Disaggregation for MLLM Inference in Edge-Cloud Networks"],"prefix":"10.1109","volume":"25","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-2176-3931","authenticated-orcid":false,"given":"Hualong","family":"Huang","sequence":"first","affiliation":[{"name":"School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1851-7185","authenticated-orcid":false,"given":"Wenhan","family":"Zhan","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7721-7422","authenticated-orcid":false,"given":"Hancong","family":"Duan","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4809-2234","authenticated-orcid":false,"given":"Kai","family":"Peng","sequence":"additional","affiliation":[{"name":"College of Engineering, Huaqiao University, Quanzhou, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1395-7314","authenticated-orcid":false,"given":"Geyong","family":"Min","sequence":"additional","affiliation":[{"name":"Department of Computer Science, University of Exeter, Exeter, U.K."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-8242-6092","authenticated-orcid":false,"given":"Zijia","family":"Zhao","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8605-0938","authenticated-orcid":false,"given":"Zitian","family":"Zhao","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yalan","family":"Ye","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1145\/3719664"},{"key":"ref2","first-page":"19274","article-title":"Fast inference from transformers via speculative decoding","volume-title":"Proc. 40th Int. Conf. Mach. Learn.","volume":"202","author":"Leviathan","year":"2023"},{"key":"ref3","first-page":"28935","article-title":"EAGLE: Speculative sampling requires rethinking feature uncertainty","volume-title":"Proc. 41st Int. Conf. Mach. Learn.","author":"Li","year":"2024"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-emnlp.767"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3620666.3651335"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/JSAC.2025.3574617"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/COMST.2025.3527641"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/JIOT.2024.3524255"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/TMC.2024.3513457"},{"key":"ref10","first-page":"1","article-title":"An image is worth 16 \u00d7 16 words: Transformers for image recognition at scale","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Dosovitskiy","year":"2021"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref12","first-page":"13937","article-title":"DynamicViT: Efficient vision transformers with dynamic token sparsification","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Rao","year":"2021"},{"key":"ref13","first-page":"1","article-title":"EVit: Expediting vision transformers via token reorganizations","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Liang","year":"2022"},{"key":"ref14","first-page":"1","article-title":"Token merging: Your vit but faster","volume-title":"Proc. 11th Int. Conf. Learn. Representations","author":"Bolya","year":"2023"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73004-7_2"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-emnlp.235"},{"key":"ref17","first-page":"1","article-title":"EAGLE-3: Scaling up inference acceleration of large language models via training-time test","volume-title":"Proc. Annu. Conf. Neural Inf. Process. Syst.","author":"Li","year":"2025"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/3676536.3676695"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v39i23.34690"},{"key":"ref20","article-title":"Recurrent drafter for fast speculative decoding in large language models","author":"Zhang","year":"2024"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/TMC.2024.3415661"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/TWC.2024.3497923"},{"key":"ref23","article-title":"QWEN-VL: A frontier large vision-language model with versatile abilities","author":"Bai","year":"2023"},{"key":"ref24","first-page":"193","article-title":"DistServe: Disaggregating prefill and decoding for goodput-optimized large language model serving","volume-title":"Proc. 18th USENIX Symp. Operating Syst. Des. Implementation","author":"Zhong","year":"2024"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1016\/j.peva.2007.06.012"},{"key":"ref26","first-page":"12","article-title":"Token pooling in vision transformers","volume-title":"Proc. 2023 IEEE\/CVF Winter Conf. Appl. Comput. Vis.","author":"Marin","year":"2023"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1109\/WACV57701.2024.00141"},{"key":"ref29","first-page":"5209","article-title":"MEDUSA: Simple LLM inference acceleration framework with multiple decoding heads","volume-title":"Proc. 41st Int. Conf. Mach. Learn.","author":"Cai","year":"2024"},{"key":"ref30","first-page":"3104","article-title":"Sequence to sequence learning with neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"27","author":"Sutskever","year":"2014"},{"key":"ref31","article-title":"The wonder shaper 1.4.1","author":"Hubert","year":"2025"},{"key":"ref32","article-title":"Expanding performance boundaries of open-source multimodal models with model, data, and test-time scaling","author":"Chen","year":"2024"},{"key":"ref33","article-title":"Llava-onevision: Easy visual task transfer","author":"Li","year":"2024"},{"key":"ref34","article-title":"hugging face\u2013the AI community building the future","author":"Face","year":"2025"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/WACV48630.2021.00225"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.findings-acl.177"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/WACV51458.2022.00264"},{"key":"ref38","first-page":"216","article-title":"MMBench: Is your multi-modal model an all-around player?","volume-title":"Proc. Proc. Eur. Conf. Comput. Vis.","author":"Liu","year":"2024"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00913"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.00868"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.52202\/079017-4116"}],"container-title":["IEEE Transactions on Mobile Computing"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/7755\/11595771\/11409366.pdf?arnumber=11409366","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T05:34:25Z","timestamp":1783402465000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11409366\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8]]},"references-count":41,"journal-issue":{"issue":"8"},"URL":"https:\/\/doi.org\/10.1109\/tmc.2026.3667591","relation":{},"ISSN":["1536-1233","1558-0660","2161-9875"],"issn-type":[{"value":"1536-1233","type":"print"},{"value":"1558-0660","type":"electronic"},{"value":"2161-9875","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8]]}}}