{"id":"https://openalex.org/W7138165718","doi":"https://doi.org/10.1609/aaai.v40i10.37771","title":"SpecDiff: Accelerating Diffusion Model Inference with Self-Speculation","display_name":"SpecDiff: Accelerating Diffusion Model Inference with Self-Speculation","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7138165718","doi":"https://doi.org/10.1609/aaai.v40i10.37771"},"language":null,"primary_location":{"id":"doi:10.1609/aaai.v40i10.37771","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i10.37771","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.1609/aaai.v40i10.37771","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5112313376","display_name":"Jiayi Pan","orcid":"https://orcid.org/0009-0002-3432-855X"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiayi Pan","raw_affiliation_strings":["Shanghai Jiao Tong University\nInfinigence-AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University\nInfinigence-AI","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123451313","display_name":"Jiaming Xu","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]},{"id":"https://openalex.org/I4210158445","display_name":"Institute of Natural Science","ror":"https://ror.org/004tze884","country_code":"KP","type":"education","lineage":["https://openalex.org/I4210158445"]}],"countries":["CN","KP"],"is_corresponding":false,"raw_author_name":"Jiaming Xu","raw_affiliation_strings":["Shanghai Jiao Tong University\nShanghai Innovation Institute"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University\nShanghai Innovation Institute","institution_ids":["https://openalex.org/I183067930","https://openalex.org/I4210158445"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129672863","display_name":"Yongkang Zhou","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]},{"id":"https://openalex.org/I4210158445","display_name":"Institute of Natural Science","ror":"https://ror.org/004tze884","country_code":"KP","type":"education","lineage":["https://openalex.org/I4210158445"]}],"countries":["CN","KP"],"is_corresponding":false,"raw_author_name":"Yongkang Zhou","raw_affiliation_strings":["Shanghai Jiao Tong University\nShanghai Innovation Institute"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University\nShanghai Innovation Institute","institution_ids":["https://openalex.org/I183067930","https://openalex.org/I4210158445"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5129718543","display_name":"Guohao Dai","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Guohao Dai","raw_affiliation_strings":["Shanghai Jiao Tong University\nInfinigence-AI\nShanghai Innovation Institute"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University\nInfinigence-AI\nShanghai Innovation Institute","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":"40","issue":"10","first_page":"8233","last_page":"8241"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.30230000615119934,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.30230000615119934,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T13702","display_name":"Machine Learning in Healthcare","score":0.06769999861717224,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11206","display_name":"Model Reduction and Neural Networks","score":0.065700002014637,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.824400007724762},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.7402999997138977},{"id":"https://openalex.org/keywords/feature-selection","display_name":"Feature selection","score":0.6446999907493591},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5965999960899353},{"id":"https://openalex.org/keywords/cache","display_name":"Cache","score":0.5364000201225281},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.5198000073432922},{"id":"https://openalex.org/keywords/similarity","display_name":"Similarity (geometry)","score":0.4602000117301941},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.3970000147819519},{"id":"https://openalex.org/keywords/feature-model","display_name":"Feature model","score":0.38609999418258667}],"concepts":[{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.824400007724762},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7888000011444092},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.7402999997138977},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.6446999907493591},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5965999960899353},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.5364000201225281},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.5220999717712402},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.5198000073432922},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.4602000117301941},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4458000063896179},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.3970000147819519},{"id":"https://openalex.org/C101814296","wikidata":"https://www.wikidata.org/wiki/Q5439685","display_name":"Feature model","level":3,"score":0.38609999418258667},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3684999942779541},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3610000014305115},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.34529998898506165},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.33809998631477356},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.335099995136261},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.32429999113082886},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.3068000078201294},{"id":"https://openalex.org/C2780573568","wikidata":"https://www.wikidata.org/wiki/Q16743864","display_name":"Fist","level":2,"score":0.3059999942779541},{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.3012999892234802},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.30059999227523804},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.3005000054836273},{"id":"https://openalex.org/C83665646","wikidata":"https://www.wikidata.org/wiki/Q42139305","display_name":"Feature vector","level":2,"score":0.2985999882221222},{"id":"https://openalex.org/C2777851325","wikidata":"https://www.wikidata.org/wiki/Q7094102","display_name":"Online model","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C124304363","wikidata":"https://www.wikidata.org/wiki/Q673661","display_name":"Abstraction","level":2,"score":0.2921000123023987},{"id":"https://openalex.org/C137635306","wikidata":"https://www.wikidata.org/wiki/Q182667","display_name":"Pareto principle","level":2,"score":0.2870999872684479},{"id":"https://openalex.org/C2778869765","wikidata":"https://www.wikidata.org/wiki/Q6028363","display_name":"Inefficiency","level":2,"score":0.2840000092983246},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.28279998898506165},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.26089999079704285},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2565000057220459}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1609/aaai.v40i10.37771","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i10.37771","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1609/aaai.v40i10.37771","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i10.37771","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Feature":[0,120],"caching":[1,25,43,105],"has":[2],"recently":[3],"emerged":[4],"as":[5],"a":[6,71,101,108,114,129,174],"promising":[7],"method":[8],"for":[9,133],"diffusion":[10,34,232],"model":[11,233],"acceleration.":[12],"It":[13],"effectively":[14],"alleviates":[15],"the":[16,29,33,46,82,86,149,166,217,222,230],"inefficiency":[17],"problem":[18],"caused":[19],"by":[20,24,164],"high":[21],"computational":[22],"requirements":[23],"similar":[26],"features":[27],"in":[28,168,195,229],"inference":[30],"process":[31],"of":[32,48,225],"model.":[35],"In":[36],"this":[37,96],"paper,":[38],"we":[39,69,98],"analyze":[40],"existing":[41],"feature":[42,104,110,116,146,154,159,169,176],"methods":[44],"from":[45],"perspective":[47],"information":[49,59,77,83,139],"utilization,":[50],"and":[51,65,113,140,143,172,188,200,212,227],"point":[52],"out":[53],"that":[54,74,182],"relying":[55],"solely":[56],"on":[57,81,95,124,137,158,205],"historical":[58,141,213],"will":[60],"lead":[61],"to":[62,203],"constrained":[63],"accuracy":[64,228],"speed":[66],"performance.":[67],"And":[68],"propose":[70],"novel":[72],"paradigm":[73],"introduces":[75,173],"future":[76],"via":[78],"self-speculation":[79],"based":[80,123,136,157],"similarity":[84],"at":[85],"same":[87],"time":[88],"step":[89],"across":[90],"different":[91],"iteration":[92],"times.":[93],"Based":[94],"paradigm,":[97],"present":[99],"SpecDiff,":[100],"training-free":[102],"multi-level":[103,115,175],"strategy":[106],"including":[107],"cached":[109,145],"selection":[111,121,147],"algorithm":[112,122,156],"classification":[117,155],"algorithm.":[118],"(1)":[119],"self-speculative":[125,138],"information.":[126],"SpecDiff":[127,183,215],"determines":[128],"dynamic":[130],"importance":[131,150,160,170],"score":[132],"each":[134],"token":[135],"information,":[142,214],"performs":[144],"through":[148],"score.":[151],"(2)":[152],"Multi-level":[153],"scores.SpecDiff":[161],"classifies":[162],"tokens":[163],"leveraging":[165],"differences":[167],"scores":[171],"calculation":[177],"strategy.":[178],"Extensive":[179],"experiments":[180],"show":[181],"achieves":[184],"average":[185],"2.80\u00d7,":[186],"2.74\u00d7,":[187],"3.17\u00d7":[189],"speedup":[190,226],"with":[191],"negligible":[192],"quality":[193],"loss":[194],"Stable":[196],"Diffusion":[197],"3,":[198],"3.5,":[199],"FLUX":[201],"compared":[202],"RFlow":[204],"NVIDIA":[206],"A800-80GB":[207],"GPU.":[208],"By":[209],"merging":[210],"speculative":[211],"overcomes":[216],"speedup-accuracy":[218],"trade-off":[219],"bottleneck,":[220],"pushing":[221],"Pareto":[223],"frontier":[224],"efficient":[231],"inference.":[234]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-03-18T00:00:00"}