{"id":"https://openalex.org/W7214760448","doi":"https://doi.org/10.48550/arxiv.2609.31873","title":"CueKFS: Agentic Cue-Driven Keyframe Selection for Long Video Understanding","display_name":"CueKFS: Agentic Cue-Driven Keyframe Selection for Long Video Understanding","publication_year":2026,"publication_date":"2026-09-25","ids":{"openalex":"https://openalex.org/W7214760448","doi":"https://doi.org/10.48550/arxiv.2609.31873"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.31873","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.31873","pdf_url":"https://arxiv.org/pdf/2609.31873","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":null,"license_id":null,"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/2609.31873","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5113395507","display_name":"Weitai Kang","orcid":"https://orcid.org/0009-0007-6484-0665","observed_orcids":["https://orcid.org/0009-0007-6484-0665"]},"institutions":[{"id":"https://openalex.org/I39422238","display_name":"University of Illinois Chicago","ror":"https://ror.org/02mpq6x41","country_code":"US","type":"education","lineage":["https://openalex.org/I39422238"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kang, Weitai","raw_affiliation_strings":["University of Illinois Chicago"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Illinois Chicago","institution_ids":["https://openalex.org/I39422238"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083681053","display_name":"Hanieh Deilamsalehy","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I869089601","display_name":"Netflix (United States)","ror":"https://ror.org/0197qw696","country_code":"US","type":"company","lineage":["https://openalex.org/I869089601"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Deilamsalehy, Hanieh","raw_affiliation_strings":["Netflix"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Netflix","institution_ids":["https://openalex.org/I869089601"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023350949","display_name":"Yumo Xu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I869089601","display_name":"Netflix (United States)","ror":"https://ror.org/0197qw696","country_code":"US","type":"company","lineage":["https://openalex.org/I869089601"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xu, Yumo","raw_affiliation_strings":["Netflix"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Netflix","institution_ids":["https://openalex.org/I869089601"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5095886520","display_name":"Dewang Sultania","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I869089601","display_name":"Netflix (United States)","ror":"https://ror.org/0197qw696","country_code":"US","type":"company","lineage":["https://openalex.org/I869089601"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sultania, Dewang","raw_affiliation_strings":["Netflix"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Netflix","institution_ids":["https://openalex.org/I869089601"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050224651","display_name":"Serdar Cellat","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I869089601","display_name":"Netflix (United States)","ror":"https://ror.org/0197qw696","country_code":"US","type":"company","lineage":["https://openalex.org/I869089601"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Cellat, Serdar","raw_affiliation_strings":["Netflix"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Netflix","institution_ids":["https://openalex.org/I869089601"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100395035","display_name":"Yan Yan","orcid":"https://orcid.org/0000-0002-7618-119X","observed_orcids":["https://orcid.org/0000-0002-7618-119X","https://orcid.org/0000-0002-3674-7160","https://orcid.org/0009-0007-1227-8594","https://orcid.org/0000-0001-9108-6767","https://orcid.org/0000-0002-3182-1739","https://orcid.org/0000-0002-1669-0662"]},"institutions":[{"id":"https://openalex.org/I39422238","display_name":"University of Illinois Chicago","ror":"https://ror.org/02mpq6x41","country_code":"US","type":"education","lineage":["https://openalex.org/I39422238"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yan, Yan","raw_affiliation_strings":["University of Illinois Chicago"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Illinois Chicago","institution_ids":["https://openalex.org/I39422238"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.5976999998092651,"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":0.5976999998092651,"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/T11439","display_name":"Video Analysis and Summarization","score":0.392300009727478,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.0020000000949949026,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/keyframe-selection","display_name":"keyframe selection","score":0.9850000143051147},{"id":"https://openalex.org/keywords/long-video-understanding","display_name":"long video understanding","score":0.7229999899864197},{"id":"https://openalex.org/keywords/video-question-answering","display_name":"video question answering","score":0.7839999794960022},{"id":"https://openalex.org/keywords/vision-language-models","display_name":"vision-language models","score":0.699999988079071},{"id":"https://openalex.org/keywords/training-free-methods","display_name":"training-free methods","score":0.5550000071525574},{"id":"https://openalex.org/keywords/question-decomposition","display_name":"question decomposition","score":0.5889999866485596},{"id":"https://openalex.org/keywords/video-summarization","display_name":"video summarization","score":0.6570000052452087},{"id":"https://openalex.org/keywords/budget-allocation","display_name":"budget allocation","score":0.3659999966621399}],"concepts":[{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.7748000025749207},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.7688000202178955},{"id":"https://openalex.org/C2780719617","wikidata":"https://www.wikidata.org/wiki/Q1030752","display_name":"Salient","level":2,"score":0.6759999990463257},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.675599992275238},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6650000214576721},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.6388000249862671},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.6279000043869019},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5835999846458435},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.4429999887943268},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.4323999881744385},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.4259999990463257},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.3801000118255615},{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.36880001425743103},{"id":"https://openalex.org/C56461940","wikidata":"https://www.wikidata.org/wiki/Q970687","display_name":"Eye tracking","level":2,"score":0.35440000891685486},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.33379998803138733},{"id":"https://openalex.org/C143271835","wikidata":"https://www.wikidata.org/wiki/Q254515","display_name":"Similitude","level":2,"score":0.32760000228881836},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.31130000948905945},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3019999861717224},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.2969000041484833},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2955999970436096},{"id":"https://openalex.org/C2775973920","wikidata":"https://www.wikidata.org/wiki/Q3252726","display_name":"Selection algorithm","level":3,"score":0.2847000062465668},{"id":"https://openalex.org/C172849965","wikidata":"https://www.wikidata.org/wiki/Q3148875","display_name":"Reference frame","level":3,"score":0.2838999927043915},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.28349998593330383},{"id":"https://openalex.org/C142853389","wikidata":"https://www.wikidata.org/wiki/Q744778","display_name":"Association (psychology)","level":2,"score":0.27230000495910645}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.31873","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.31873","pdf_url":"https://arxiv.org/pdf/2609.31873","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.31873","doi":"https://doi.org/10.48550/arxiv.2609.31873","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.31873","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.31873","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.31873","pdf_url":"https://arxiv.org/pdf/2609.31873","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":null,"license_id":null,"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/W7214760448.pdf","grobid_xml":"https://content.openalex.org/works/W7214760448.grobid-xml"},"referenced_works_count":23,"referenced_works":["https://openalex.org/W3173181308","https://openalex.org/W4241400486","https://openalex.org/W4400376579","https://openalex.org/W4403928665","https://openalex.org/W4407794264","https://openalex.org/W4407815020","https://openalex.org/W4412888065","https://openalex.org/W4413146798","https://openalex.org/W4414384364","https://openalex.org/W4414971583","https://openalex.org/W4415798209","https://openalex.org/W4416053382","https://openalex.org/W4417035275","https://openalex.org/W4417095620","https://openalex.org/W6929565228","https://openalex.org/W6948405068","https://openalex.org/W7131825757","https://openalex.org/W7133362550","https://openalex.org/W7139917033","https://openalex.org/W7148492639","https://openalex.org/W7155052898","https://openalex.org/W7160137148","https://openalex.org/W7162339592"],"related_works":["https://openalex.org/W7167094649","https://openalex.org/W7216547966","https://openalex.org/W7167237393","https://openalex.org/W4415288380","https://openalex.org/W4417466313","https://openalex.org/W7138326392","https://openalex.org/W7214804103","https://openalex.org/W7154704948","https://openalex.org/W4415918554","https://openalex.org/W7137846659"],"abstract_inverted_index":{"Keyframe":[0],"selection":[1],"(KFS)":[2],"has":[3],"long":[4],"produced":[5],"compact":[6],"video":[7,143],"summaries":[8],"for":[9,16,32],"browsing":[10],"and":[11,13,202],"retrieval,":[12],"representative":[14],"frames":[15,38,49,116],"thumbnails.":[17],"More":[18],"recently,":[19],"when":[20,62],"conditioned":[21],"on":[22],"a":[23,56,107,118,203,213],"question,":[24],"KFS":[25],"provides":[26],"an":[27,126],"alternative":[28],"to":[29,43,52,85,98,144,146,162,196,236],"uniform":[30],"sampling":[31],"long-video":[33],"question":[34,64,89,135],"answering":[35],"by":[36,50],"selecting":[37],"that":[39,69,110,220],"are":[40],"more":[41],"relevant":[42,57],"the":[44,53,63,88,134,142,156,164,169,172,199,228,239],"question.":[45,54],"Most":[46],"methods":[47,83],"rank":[48],"similarity":[51,232],"Yet":[55],"frame":[58,72],"may":[59],"score":[60],"poorly":[61],"combines":[65],"subjects":[66],"or":[67,74],"moments":[68],"no":[70],"single":[71],"shows,":[73],"requires":[75],"implicit":[76],"information":[77],"absent":[78],"from":[79,94],"its":[80,147,160],"wording.":[81],"Other":[82],"try":[84],"break":[86],"down":[87],"into":[90,136],"subqueries,":[91],"but":[92],"suffer":[93],"inaccurate":[95],"decomposition":[96],"due":[97],"their":[99],"static":[100],"initial":[101,127,240],"context.":[102,241],"Therefore,":[103],"we":[104,132],"propose":[105],"CueKFS,":[106,218],"training-free":[108],"method":[109],"reformulates":[111],"question--frame":[112],"matching":[113],"as":[114],"comparing":[115],"against":[117,159],"set":[119,128,158],"of":[120,129,194,205,217,227,234],"dynamically":[121],"generated":[122],"visual":[123],"cues.":[124,137,174],"From":[125],"salient":[130],"frames,":[131],"decompose":[133],"Each":[138],"cue":[139,157,222],"concurrently":[140],"probes":[141],"navigate":[145],"own":[148],"evidence.":[149],"A":[150],"reasoning":[151],"VLM":[152,208],"then":[153,167],"agentically":[154],"revises":[155],"evidence":[161],"re-explore":[163],"video.":[165],"CueKFS":[166,178],"allocates":[168],"budget":[170],"across":[171],"surviving":[173],"Across":[175],"three":[176],"benchmarks,":[177],"establishes":[179],"state-of-the-art":[180],"results":[181],"in":[182],"all":[183],"27":[184],"evaluated":[185],"settings":[186],"with":[187],"available":[188],"prior":[189],"results,":[190],"achieving":[191],"budget-averaged":[192],"gains":[193,233],"up":[195,235],"+4.54%":[197],"over":[198,238],"previous":[200],"baseline":[201],"median":[204],"only":[206],"two":[207],"calls.":[209],"We":[210],"further":[211],"provide":[212],"detailed":[214],"behavioral":[215],"analysis":[216],"showing":[219],"agentic":[221],"refinement":[223],"drives":[224],"active":[225],"re-exploration":[226],"video,":[229],"yielding":[230],"relative":[231],"92%":[237]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-30T00:00:00"}