{"id":"https://openalex.org/W7153194163","doi":"https://doi.org/10.48550/arxiv.2604.08077","title":"AdaSpark: Adaptive Sparsity for Efficient Long-Video Understanding","display_name":"AdaSpark: Adaptive Sparsity for Efficient Long-Video Understanding","publication_year":2026,"publication_date":"2026-04-09","ids":{"openalex":"https://openalex.org/W7153194163","doi":"https://doi.org/10.48550/arxiv.2604.08077"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.08077","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08077","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.08077","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133380320","display_name":"Handong Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Handong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133333729","display_name":"Zikang Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zikang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133338600","display_name":"Longteng Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Longteng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101322148","display_name":"Tongtian Yue","orcid":"https://orcid.org/0000-0001-5774-4084"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yue, Tongtian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133383687","display_name":"Yepeng Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Yepeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054313662","display_name":"Xinxin Zhu","orcid":"https://orcid.org/0000-0001-6229-5615"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Xinxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133333923","display_name":"Chuanyang Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Chuanyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133380948","display_name":"Ziming Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Ziming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133362002","display_name":"Zhibin Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zhibin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133322493","display_name":"Jun Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Jun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133345196","display_name":"Cheng Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Cheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133362815","display_name":"Bo Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Bo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100374980","display_name":"Jing Liu","orcid":"https://orcid.org/0000-0002-2235-3961"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Jing","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":false,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.4715999960899353,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.4715999960899353,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.23309999704360962,"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/T10812","display_name":"Human Pose and Action Recognition","score":0.08030000329017639,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/salient","display_name":"Salient","score":0.5679000020027161},{"id":"https://openalex.org/keywords/flops","display_name":"FLOPS","score":0.5217000246047974},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.5162000060081482},{"id":"https://openalex.org/keywords/perception","display_name":"Perception","score":0.3910999894142151},{"id":"https://openalex.org/keywords/computational-complexity-theory","display_name":"Computational complexity theory","score":0.37209999561309814},{"id":"https://openalex.org/keywords/identification","display_name":"Identification (biology)","score":0.3253999948501587}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.807200014591217},{"id":"https://openalex.org/C2780719617","wikidata":"https://www.wikidata.org/wiki/Q1030752","display_name":"Salient","level":2,"score":0.5679000020027161},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.5217000246047974},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.5162000060081482},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5040000081062317},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.3910999894142151},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.37209999561309814},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36239999532699585},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.3253999948501587},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.287200003862381},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.28130000829696655},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.272599995136261},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.27250000834465027},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2718999981880188},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.26899999380111694},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.2599000036716461},{"id":"https://openalex.org/C2776003309","wikidata":"https://www.wikidata.org/wiki/Q1988072","display_name":"Adaptive algorithm","level":2,"score":0.25679999589920044}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.08077","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08077","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.08077","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08077","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"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":{"Processing":[0],"long-form":[1],"videos":[2],"with":[3],"Video":[4],"Large":[5],"Language":[6],"Models":[7],"(Video-LLMs)":[8],"is":[9],"computationally":[10],"prohibitive.":[11],"Current":[12],"efficiency":[13],"methods":[14],"often":[15],"compromise":[16],"fine-grained":[17],"perception":[18],"through":[19],"irreversible":[20],"information":[21],"disposal":[22],"or":[23],"inhibit":[24],"long-range":[25,135],"temporal":[26],"modeling":[27],"via":[28],"rigid,":[29],"predefined":[30],"sparse":[31],"patterns.":[32],"This":[33],"paper":[34],"introduces":[35],"AdaSpark,":[36],"an":[37],"adaptive":[38],"sparsity":[39],"framework":[40],"designed":[41],"to":[42,76,122,129],"address":[43],"these":[44],"limitations.":[45],"AdaSpark":[46,115],"first":[47],"partitions":[48],"video":[49,74,142],"inputs":[50],"into":[51],"3D":[52],"spatio-temporal":[53],"cubes.":[54],"It":[55],"then":[56],"employs":[57],"two":[58],"co-designed,":[59],"context-aware":[60],"components:":[61],"(1)":[62],"Adaptive":[63,84],"Cube-Selective":[64],"Attention":[65],"(AdaS-Attn),":[66],"which":[67,88],"adaptively":[68,104],"selects":[69],"a":[70],"subset":[71],"of":[72],"relevant":[73],"cubes":[75],"attend":[77],"for":[78],"each":[79,97],"query":[80],"token,":[81],"and":[82,132],"(2)":[83],"Token-Selective":[85],"FFN":[86],"(AdaS-FFN),":[87],"selectively":[89],"processes":[90],"only":[91],"the":[92],"most":[93],"salient":[94],"tokens":[95],"within":[96],"cube.":[98],"An":[99],"entropy-based":[100],"(Top-p)":[101],"selection":[102],"mechanism":[103],"allocates":[105],"computational":[106,118],"resources":[107],"based":[108],"on":[109,139],"input":[110],"complexity.":[111],"Experiments":[112],"demonstrate":[113],"that":[114],"significantly":[116],"reduces":[117],"load":[119],"by":[120],"up":[121],"57%":[123],"FLOPs":[124],"while":[125],"maintaining":[126],"comparable":[127],"performance":[128],"dense":[130],"models":[131],"preserving":[133],"fine-grained,":[134],"dependencies,":[136],"as":[137],"validated":[138],"challenging":[140],"hour-scale":[141],"benchmarks.":[143]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-04-11T00:00:00"}