{"id":"https://openalex.org/W7138352668","doi":"https://doi.org/10.1609/aaai.v40i4.37287","title":"Sparse-vDiT: Unleashing the Power of Sparse Attention to Accelerate Video Diffusion Transformers","display_name":"Sparse-vDiT: Unleashing the Power of Sparse Attention to Accelerate Video Diffusion Transformers","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7138352668","doi":"https://doi.org/10.1609/aaai.v40i4.37287"},"language":null,"primary_location":{"id":"doi:10.1609/aaai.v40i4.37287","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i4.37287","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/37287/41249","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://ojs.aaai.org/index.php/AAAI/article/download/37287/41249","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129726128","display_name":"Pengtao Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Pengtao Chen","raw_affiliation_strings":["Fudan University\nStepFun"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Fudan University\nStepFun","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129665608","display_name":"Xianfang Zeng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xianfang Zeng","raw_affiliation_strings":["StepFun"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"StepFun","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5079198841","display_name":"Mengfei Zhao","orcid":"https://orcid.org/0009-0005-4476-8657"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Maosen Zhao","raw_affiliation_strings":["Fudan University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Fudan University","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073283125","display_name":"Mingzhu Shen","orcid":"https://orcid.org/0000-0003-1357-556X"},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Mingzhu Shen","raw_affiliation_strings":["Imperial College London"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Imperial College London","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129739881","display_name":"Wei Cheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Cheng","raw_affiliation_strings":["StepFun"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"StepFun","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129688447","display_name":"Gang Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gang Yu","raw_affiliation_strings":["StepFun"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"StepFun","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5129643777","display_name":"Tao Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Tao Chen","raw_affiliation_strings":["Fudan University\nShanghai Innovation Institute"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Fudan University\nShanghai Innovation Institute","institution_ids":["https://openalex.org/I24943067"]}]}],"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":true,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":"40","issue":"4","first_page":"2957","last_page":"2965"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.3977000117301941,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.3977000117301941,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.06210000067949295,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10623","display_name":"Thin-Film Transistor Technologies","score":0.05590000003576279,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6450999975204468},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5705999732017517},{"id":"https://openalex.org/keywords/quadratic-equation","display_name":"Quadratic equation","score":0.5443000197410583},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.48069998621940613},{"id":"https://openalex.org/keywords/flops","display_name":"FLOPS","score":0.42579999566078186},{"id":"https://openalex.org/keywords/fuse","display_name":"Fuse (electrical)","score":0.4056999981403351},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.3594000041484833},{"id":"https://openalex.org/keywords/quadratic-programming","display_name":"Quadratic programming","score":0.3449000120162964},{"id":"https://openalex.org/keywords/visualization","display_name":"Visualization","score":0.3140000104904175}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7390000224113464},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6450999975204468},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5705999732017517},{"id":"https://openalex.org/C129844170","wikidata":"https://www.wikidata.org/wiki/Q41299","display_name":"Quadratic equation","level":2,"score":0.5443000197410583},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.48069998621940613},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.42579999566078186},{"id":"https://openalex.org/C141353440","wikidata":"https://www.wikidata.org/wiki/Q182221","display_name":"Fuse (electrical)","level":2,"score":0.4056999981403351},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.39809998869895935},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3944999873638153},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.3594000041484833},{"id":"https://openalex.org/C81845259","wikidata":"https://www.wikidata.org/wiki/Q290117","display_name":"Quadratic programming","level":2,"score":0.3449000120162964},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3140000104904175},{"id":"https://openalex.org/C124066611","wikidata":"https://www.wikidata.org/wiki/Q28684319","display_name":"Sparse approximation","level":2,"score":0.3068000078201294},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.30160000920295715},{"id":"https://openalex.org/C56372850","wikidata":"https://www.wikidata.org/wiki/Q1050404","display_name":"Sparse matrix","level":3,"score":0.30090001225471497},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.29899999499320984},{"id":"https://openalex.org/C65483669","wikidata":"https://www.wikidata.org/wiki/Q3536669","display_name":"Video processing","level":2,"score":0.2930999994277954},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.29159998893737793},{"id":"https://openalex.org/C186967261","wikidata":"https://www.wikidata.org/wiki/Q5082128","display_name":"Mobile device","level":2,"score":0.2849999964237213},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.28369998931884766},{"id":"https://openalex.org/C26713055","wikidata":"https://www.wikidata.org/wiki/Q245962","display_name":"Implementation","level":2,"score":0.2750999927520752},{"id":"https://openalex.org/C39482219","wikidata":"https://www.wikidata.org/wiki/Q192826","display_name":"Kronecker delta","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2660999894142151},{"id":"https://openalex.org/C49020025","wikidata":"https://www.wikidata.org/wiki/Q1059099","display_name":"Chaining","level":2,"score":0.2565999925136566},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.2538999915122986},{"id":"https://openalex.org/C2777472644","wikidata":"https://www.wikidata.org/wiki/Q16968992","display_name":"Approximate inference","level":3,"score":0.25110000371932983}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1609/aaai.v40i4.37287","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i4.37287","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/37287/41249","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.v40i4.37287","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i4.37287","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/37287/41249","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":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7138352668.pdf","grobid_xml":"https://content.openalex.org/works/W7138352668.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"While":[0],"Diffusion":[1,37],"Transformers":[2],"(DiTs)":[3],"have":[4],"achieved":[5],"breakthroughs":[6],"in":[7,25,35,197],"video":[8,205],"generation,":[9],"this":[10],"long":[11,204],"sequence":[12],"generation":[13],"task":[14],"remains":[15],"constrained":[16],"by":[17],"the":[18,73,115,130,137,142],"quadratic":[19],"complexity":[20],"of":[21,32,171],"attention":[22,33,54,96,144],"mechanisms,":[23],"resulting":[24],"significant":[26],"inference":[27,147,169],"latency.":[28],"Through":[29],"detailed":[30],"analysis":[31],"maps":[34],"Video":[36],"Transformer":[38],"(vDiT),":[39],"we":[40,79,133],"identify":[41],"three":[42],"recurring":[43],"sparsity":[44,83,104,196],"patterns:":[45],"diagonal,":[46],"multi-diagonal,":[47],"and":[48,65,122,156,162,167,174,188],"vertical-stripe":[49],"structures.":[50],"And":[51],"even":[52],"3-6%":[53],"heads":[55,135],"can":[56,199],"be":[57,200],"skipped.":[58],"Crucially,":[59],"these":[60,77],"patterns":[61],"exhibit":[62],"strong":[63],"layer-depth":[64],"head-position":[66],"correlations":[67],"but":[68],"show":[69],"limited":[70],"dependence":[71],"on":[72],"input":[74],"content.":[75],"Leveraging":[76],"findings,":[78],"propose":[80],"Sparse-vDiT,":[81],"a":[82],"acceleration":[84],"framework":[85],"for":[86,101,203],"vDiT":[87,152],"comprising:":[88],"1)":[89],"Pattern-optimized":[90],"sparse":[91,109,117],"kernels":[92],"that":[93,113,140,193],"replace":[94],"dense":[95],"with":[97,182],"computationally":[98],"efficient":[99],"implementations":[100],"each":[102],"identified":[103],"pattern.":[105],"2)":[106],"An":[107],"offline":[108],"diffusion":[110],"search":[111],"algorithm":[112],"selects":[114],"optimal":[116,131],"computation":[118],"strategy":[119],"per":[120],"layer":[121,139],"head":[123],"via":[124],"hardware-aware":[125],"cost":[126],"modeling.":[127],"After":[128],"determining":[129],"configuration,":[132],"fuse":[134],"within":[136],"same":[138,143],"share":[141],"strategy,":[145],"enhancing":[146],"efficiency.":[148],"Integrated":[149],"into":[150],"state-of-the-art":[151],"models":[153],"(CogVideoX1.5,":[154],"HunyuanVideo,":[155],"Wan2.1),":[157],"Sparse-vDiT":[158],"achieves":[159],"2.09\u00d7,":[160],"2.38\u00d7,":[161],"1.67\u00d7":[163],"theoretical":[164],"FLOP":[165],"reduction,":[166],"actual":[168],"speedups":[170],"1.76\u00d7,":[172],"1.85\u00d7,":[173],"1.58\u00d7,":[175],"respectively,":[176],"while":[177],"maintaining":[178],"high":[179],"visual":[180],"fidelity,":[181],"PSNR":[183],"values":[184],"reaching":[185],"24.13,":[186],"27.09,":[187],"22.59.":[189],"Our":[190],"work":[191],"demonstrates":[192],"latent":[194],"structural":[195],"vDiTs":[198],"systematically":[201],"exploited":[202],"synthesis.":[206]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-03-18T00:00:00"}