{"id":"https://openalex.org/W7204097809","doi":"https://doi.org/10.48550/arxiv.2608.20814","title":"Enhancing Localized Reasoning for Long Video Understanding via Efficient Segment-to-Video Supervision","display_name":"Enhancing Localized Reasoning for Long Video Understanding via Efficient Segment-to-Video Supervision","publication_year":2026,"publication_date":"2026-08-21","ids":{"openalex":"https://openalex.org/W7204097809","doi":"https://doi.org/10.48550/arxiv.2608.20814"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.20814","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.20814","pdf_url":"https://arxiv.org/pdf/2608.20814","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.20814","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5147954148","display_name":"Beibei Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405270328","display_name":"Ant Group (China)","ror":"https://ror.org/00g5w7n47","country_code":"CN","type":"company","lineage":["https://openalex.org/I4405270328"]},{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Beibei","raw_affiliation_strings":["Ant Group","State Key Laboratory for Novel Software Technology , Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group","institution_ids":["https://openalex.org/I4405270328"]},{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology , Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5143423815","display_name":"Chao Xu","orcid":"https://orcid.org/0000-0002-2759-6364","observed_orcids":["https://orcid.org/0000-0002-2759-6364"]},"institutions":[{"id":"https://openalex.org/I4405270328","display_name":"Ant Group (China)","ror":"https://ror.org/00g5w7n47","country_code":"CN","type":"company","lineage":["https://openalex.org/I4405270328"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xu, Chao","raw_affiliation_strings":["Ant Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group","institution_ids":["https://openalex.org/I4405270328"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125405144","display_name":"Jun Lan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405270328","display_name":"Ant Group (China)","ror":"https://ror.org/00g5w7n47","country_code":"CN","type":"company","lineage":["https://openalex.org/I4405270328"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lan, Jun","raw_affiliation_strings":["Ant Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group","institution_ids":["https://openalex.org/I4405270328"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147995320","display_name":"Zongyi Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405270328","display_name":"Ant Group (China)","ror":"https://ror.org/00g5w7n47","country_code":"CN","type":"company","lineage":["https://openalex.org/I4405270328"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Li, Zongyi","raw_affiliation_strings":["Ant Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group","institution_ids":["https://openalex.org/I4405270328"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137519842","display_name":"Lai Wei","orcid":null,"observed_orcids":[]},"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":"Wei, Lai","raw_affiliation_strings":["School of Computer Science , Shanghai Jiao Tong University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science , Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101236538","display_name":"Huijia Zhu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405270328","display_name":"Ant Group (China)","ror":"https://ror.org/00g5w7n47","country_code":"CN","type":"company","lineage":["https://openalex.org/I4405270328"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhu, Huijia","raw_affiliation_strings":["Ant Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Ant Group","institution_ids":["https://openalex.org/I4405270328"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5084591510","display_name":"Tongwei Ren","orcid":"https://orcid.org/0000-0003-3092-424X","observed_orcids":["https://orcid.org/0000-0003-3092-424X"]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ren, Tongwei","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology , Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology , Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"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":1.0,"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":1.0,"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":0.0,"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/T10028","display_name":"Topic Modeling","score":0.0,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/long-video-understanding","display_name":"long video understanding","score":0.9980000257492065},{"id":"https://openalex.org/keywords/multimodal-large-language-models","display_name":"multimodal large language models","score":0.8870000243186951},{"id":"https://openalex.org/keywords/video-question-answering","display_name":"video question answering","score":0.8550000190734863},{"id":"https://openalex.org/keywords/reinforcement-learning-fine-tuning","display_name":"reinforcement learning fine-tuning","score":0.7720000147819519},{"id":"https://openalex.org/keywords/training-efficiency","display_name":"training efficiency","score":0.5659999847412109},{"id":"https://openalex.org/keywords/inference-latency","display_name":"inference latency","score":0.8610000014305115}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8120999932289124},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7459999918937683},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.671999990940094},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6173999905586243},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5598000288009644},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.492900013923645},{"id":"https://openalex.org/C2779913896","wikidata":"https://www.wikidata.org/wiki/Q7063001","display_name":"Notice","level":2,"score":0.4781999886035919},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.39800000190734863},{"id":"https://openalex.org/C2776321320","wikidata":"https://www.wikidata.org/wiki/Q857525","display_name":"Annotation","level":2,"score":0.38190001249313354},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.350600004196167},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.31279999017715454},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.31029999256134033},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.30320000648498535},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.2992999851703644},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.2962000072002411},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2590000033378601}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.20814","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.20814","pdf_url":"https://arxiv.org/pdf/2608.20814","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.20814","doi":"https://doi.org/10.48550/arxiv.2608.20814","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.20814","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:2608.20814","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.20814","pdf_url":"https://arxiv.org/pdf/2608.20814","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":"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/W7204097809.pdf","grobid_xml":"https://content.openalex.org/works/W7204097809.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4415101752","https://openalex.org/W7201848394","https://openalex.org/W4403725941","https://openalex.org/W7129109165","https://openalex.org/W4415003439","https://openalex.org/W7168410370","https://openalex.org/W7140159845","https://openalex.org/W7124746758","https://openalex.org/W7092298218","https://openalex.org/W7117749048"],"abstract_inverted_index":{"Though":[0],"Multimodal":[1],"Large":[2],"Language":[3],"Models":[4],"(MLLMs)":[5],"have":[6,53],"shown":[7],"impressive":[8],"potential":[9],"in":[10,22,83,110,175,237,242],"video":[11,14,133],"understanding,":[12],"long":[13],"understanding":[15],"(LVU)":[16],"remains":[17],"challenging":[18],"since":[19],"distracting":[20,173],"noise":[21,174],"complex":[23],"and":[24,73,80,89,123,198,232,244],"lengthy":[25,87],"contexts":[26],"can":[27,144,161,219],"obscure":[28],"localized":[29,121],"details,":[30],"misleading":[31],"MLLMs":[32,163,231],"to":[33,46,105,130,137,150,164],"produce":[34],"incorrect":[35],"answers.":[36],"Recent":[37],"works":[38],"mitigate":[39],"these":[40,51,95,126],"issues":[41],"by":[42],"incentivizing":[43],"deep":[44],"reasoning":[45,109],"include":[47],"relevant":[48],"evidence.":[49],"However,":[50],"methods":[52,85,234],"two":[54],"main":[55],"problems:":[56],"First,":[57],"the":[58,78,131,176,199],"reinforcement":[59,184],"fine-tuning":[60],"framework":[61],"(RFT)":[62],"they":[63],"leveraged":[64],"incurs":[65],"substantial":[66],"training":[67,181,243],"overheads,":[68],"including":[69],"high":[70,90],"annotation":[71],"costs":[72],"complicated":[74],"reward":[75,191],"designs.":[76],"Second,":[77],"self-reflective":[79],"iterative-perception":[81],"mechanism":[82],"some":[84],"causes":[86],"outputs":[88],"inference":[91,245],"latency.":[92],"To":[93],"alleviate":[94],"problems,":[96],"we":[97,113],"propose":[98],"a":[99,154,188,206],"novel":[100],"Segment-to-Video":[101],"Supervision}":[102],"method":[103],"(S2V)":[104],"efficiently":[106],"enhance":[107],"fine-grained":[108,167],"LVU.":[111],"Specifically,":[112],"generate":[114],"question":[115],"answer":[116,204],"pairs":[117],"(VQA)":[118],"based":[119,192],"on":[120,139,158,193],"segments,":[122,141],"then":[124],"transfer":[125],"segment-based":[127,142],"VQA":[128,143,196],"back":[129],"whole":[132,177],"for":[134],"training.":[135],"Due":[136],"focusing":[138],"short":[140],"naturally":[145],"notice":[146],"details":[147,168],"which":[148],"tend":[149],"be":[151],"overlooked":[152],"from":[153],"whole-video":[155],"perspective.":[156],"Training":[157],"such":[159],"data":[160],"enforce":[162],"correctly":[165],"associate":[166],"with":[169,187,210],"QA":[170],"while":[171],"avoiding":[172],"video.":[178],"The":[179],"S2V":[180,201,218],"involves":[182],"just":[183],"learning":[185],"(RL)":[186],"simple":[189],"accuracy":[190,239],"only":[194,236],"10K":[195],"samples":[197],"resulting":[200],"model":[202],"predicts":[203],"using":[205],"single":[207],"forward":[208],"pass":[209],"limited":[211],"output":[212],"tokens.":[213],"Experimental":[214],"results":[215],"demonstrate":[216],"that":[217],"consistently":[220],"improve":[221],"LVU":[222,226,238],"performance":[223],"across":[224],"multiple":[225],"benchmarks,":[227],"outperforming":[228],"both":[229],"general":[230],"reasoning-based":[233],"not":[235],"but":[240],"also":[241],"efficiency.":[246]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-08-25T00:00:00"}