{"id":"https://openalex.org/W7214972709","doi":"https://doi.org/10.48550/arxiv.2609.37001","title":"Parameterized Stripe Attention for Efficient Video Generation","display_name":"Parameterized Stripe Attention for Efficient Video Generation","publication_year":2026,"publication_date":"2026-09-29","ids":{"openalex":"https://openalex.org/W7214972709","doi":"https://doi.org/10.48550/arxiv.2609.37001"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.37001","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.37001","pdf_url":"https://arxiv.org/pdf/2609.37001","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2609.37001","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5153612506","display_name":"Xingyu Jia","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jia, Xingyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5035198517","display_name":"Baole Ai","orcid":"https://orcid.org/0009-0006-2497-7328","observed_orcids":["https://orcid.org/0009-0006-2497-7328"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ai, Baole","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153573736","display_name":"Ang Wang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Ang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153603956","display_name":"Kang Zhao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Kang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5153582836","display_name":"Yong Li","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"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":0.8490999937057495,"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":0.8490999937057495,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.13259999454021454,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.005499999970197678,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/sparse-attention","display_name":"sparse attention","score":0.746999979019165},{"id":"https://openalex.org/keywords/video-diffusion-models","display_name":"video diffusion models","score":0.6330000162124634},{"id":"https://openalex.org/keywords/diffusion-transformer","display_name":"diffusion transformer","score":0.7170000076293945},{"id":"https://openalex.org/keywords/inference-latency","display_name":"inference latency","score":0.6100000143051147},{"id":"https://openalex.org/keywords/cuda-kernels","display_name":"CUDA kernels","score":0.5630000233650208}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8303999900817871},{"id":"https://openalex.org/C165464430","wikidata":"https://www.wikidata.org/wiki/Q1570441","display_name":"Parameterized complexity","level":2,"score":0.6604999899864197},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6226000189781189},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.48010000586509705},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.46639999747276306},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.45419999957084656},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.44999998807907104},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.42739999294281006},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.42160001397132874},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.4147999882698059},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.39750000834465027},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3910999894142151},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.3889000117778778},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3682999908924103},{"id":"https://openalex.org/C130367717","wikidata":"https://www.wikidata.org/wiki/Q189791","display_name":"Diagonal","level":2,"score":0.365200012922287},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.36250001192092896},{"id":"https://openalex.org/C124066611","wikidata":"https://www.wikidata.org/wiki/Q28684319","display_name":"Sparse approximation","level":2,"score":0.3199000060558319},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.2971000075340271},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2840999960899353},{"id":"https://openalex.org/C163173736","wikidata":"https://www.wikidata.org/wiki/Q3308558","display_name":"Key generation","level":3,"score":0.275299996137619},{"id":"https://openalex.org/C186967261","wikidata":"https://www.wikidata.org/wiki/Q5082128","display_name":"Mobile device","level":2,"score":0.26840001344680786},{"id":"https://openalex.org/C2776449333","wikidata":"https://www.wikidata.org/wiki/Q7928781","display_name":"View synthesis","level":3,"score":0.26460000872612},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.2597000002861023},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.25949999690055847},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.2590000033378601},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.25619998574256897},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.2533999979496002},{"id":"https://openalex.org/C2779696439","wikidata":"https://www.wikidata.org/wiki/Q7512811","display_name":"Signature (topology)","level":2,"score":0.2529999911785126},{"id":"https://openalex.org/C2777472644","wikidata":"https://www.wikidata.org/wiki/Q16968992","display_name":"Approximate inference","level":3,"score":0.25279998779296875}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.37001","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.37001","pdf_url":"https://arxiv.org/pdf/2609.37001","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.37001","doi":"https://doi.org/10.48550/arxiv.2609.37001","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.37001","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.37001","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.37001","pdf_url":"https://arxiv.org/pdf/2609.37001","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214972709.pdf","grobid_xml":"https://content.openalex.org/works/W7214972709.grobid-xml"},"referenced_works_count":20,"referenced_works":["https://openalex.org/W3092442149","https://openalex.org/W4303647933","https://openalex.org/W4399452094","https://openalex.org/W4402387301","https://openalex.org/W4404313951","https://openalex.org/W4405096483","https://openalex.org/W4405956131","https://openalex.org/W4405956418","https://openalex.org/W4406032994","https://openalex.org/W4407184867","https://openalex.org/W4407308859","https://openalex.org/W4407764033","https://openalex.org/W4409092337","https://openalex.org/W4412945604","https://openalex.org/W4414581692","https://openalex.org/W4416522910","https://openalex.org/W6892034596","https://openalex.org/W7138352668","https://openalex.org/W7160028311","https://openalex.org/W7161937071"],"related_works":["https://openalex.org/W4407184867","https://openalex.org/W4407308859","https://openalex.org/W7222852351","https://openalex.org/W6892004055","https://openalex.org/W4415135912","https://openalex.org/W7162341135","https://openalex.org/W4415320270","https://openalex.org/W7155413133","https://openalex.org/W4407386433","https://openalex.org/W7154483853"],"abstract_inverted_index":{"Diffusion":[0],"Transformers":[1],"(DiTs)":[2],"enable":[3],"high-quality":[4],"video":[5,77],"generation":[6],"but":[7],"suffer":[8],"from":[9],"substantial":[10],"inference":[11],"latency,":[12],"primarily":[13],"attributable":[14],"to":[15,41,133],"the":[16,39,57,112],"computationally":[17],"expensive":[18],"full":[19],"spatio-temporal":[20],"attention.":[21],"While":[22],"sparse":[23,136],"attention":[24,44,66,79,109,118,166],"methods":[25],"offer":[26],"potential":[27],"solutions,":[28],"existing":[29,72],"approaches":[30],"face":[31],"an":[32],"inherent":[33],"flexibility--efficiency":[34],"dilemma:":[35],"predefined":[36],"masks":[37,48],"lack":[38,58],"flexibility":[40],"capture":[42],"diverse":[43,117],"patterns,":[45,137],"while":[46],"runtime-determined":[47],"introduce":[49],"overheads":[50],"and":[51,74,88,171,178],"sacrifice":[52],"hardware":[53],"efficiency.":[54],"We":[55],"identify":[56],"of":[59,64,71],"a":[60,68,98,106,128,150,160],"unified":[61,125],"structural":[62],"characterization":[63],"DiT":[65,78],"as":[67],"key":[69],"limitation":[70],"methods,":[73],"establish":[75],"that":[76,110,155,174],"exhibits":[80],"\\textbf{periodic":[81],"diagonal":[82],"stripe":[83,108,114],"structures}":[84],"along":[85],"both":[86],"temporal":[87],"spatial":[89],"dimensions.":[90],"To":[91,143],"formally":[92],"encode":[93],"these":[94],"structured":[95],"patterns":[96,119],"within":[97],"single":[99,129],"efficient":[100,121],"kernel,":[101],"we":[102,148],"present":[103],"{\\bf":[104],"PSA},":[105],"parameterized":[107],"formalizes":[111],"observed":[113],"regularity,":[115],"unifying":[116],"for":[120,164],"mask":[122],"generation.":[123],"This":[124],"representation":[126],"enables":[127],"hardware-efficient":[130],"CUDA":[131],"kernel":[132],"process":[134],"all":[135],"achieving":[138],"FlashAttention-3-level":[139],"Model":[140],"FLOPs":[141],"Utilization.":[142],"determine":[144],"optimal":[145],"sparsity":[146,158],"configurations,":[147],"propose":[149],"training-free":[151],"offline":[152],"search":[153],"algorithm":[154],"automatically":[156],"maximizes":[157],"under":[159],"specified":[161],"error":[162],"tolerance":[163],"each":[165],"head.":[167],"Experiments":[168],"on":[169],"HunyuanVideo":[170],"Wan~2.1":[172],"demonstrate":[173],"PSA":[175],"achieves":[176],"1.57$\\times$":[177],"1.37$\\times$":[179],"end-to-end":[180],"speedups":[181],"over":[182],"FlashAttention-3":[183],"baselines,":[184],"with":[185],"acceptable":[186],"visual":[187],"quality":[188],"degradation.":[189]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-01T00:00:00"}