{"id":"https://openalex.org/W7166880001","doi":"https://doi.org/10.18653/v1/2026.acl-long.1087","title":"See the Forest for the Trees: Loosely Speculative Decoding via Visual-Semantic Guidance for Efficient Inference of Video LLMs","display_name":"See the Forest for the Trees: Loosely Speculative Decoding via Visual-Semantic Guidance for Efficient Inference of Video LLMs","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166880001","doi":"https://doi.org/10.18653/v1/2026.acl-long.1087"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-long.1087","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1087","pdf_url":"https://aclanthology.org/2026.acl-long.1087.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-long.1087.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135924110","display_name":"Yicheng Ji","orcid":null},"institutions":[{"id":"https://openalex.org/I4210156488","display_name":"Institute of Zoology of the Slovak Academy of Sciences","ror":"https://ror.org/05tm6j853","country_code":"SK","type":"facility","lineage":["https://openalex.org/I207624831","https://openalex.org/I4210156488"]},{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN","SK"],"is_corresponding":false,"raw_author_name":"Yicheng Ji","raw_affiliation_strings":["Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","The State Key Laboratory of Blockchain and Data Security , Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","institution_ids":["https://openalex.org/I4210156488"]},{"raw_affiliation_string":"The State Key Laboratory of Blockchain and Data Security , Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139834593","display_name":"Jun Zhang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210156488","display_name":"Institute of Zoology of the Slovak Academy of Sciences","ror":"https://ror.org/05tm6j853","country_code":"SK","type":"facility","lineage":["https://openalex.org/I207624831","https://openalex.org/I4210156488"]},{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN","SK"],"is_corresponding":false,"raw_author_name":"Jun Zhang","raw_affiliation_strings":["Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","The State Key Laboratory of Blockchain and Data Security , Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","institution_ids":["https://openalex.org/I4210156488"]},{"raw_affiliation_string":"The State Key Laboratory of Blockchain and Data Security , Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139819726","display_name":"Jinpeng Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jinpeng Chen","raw_affiliation_strings":["School of Computer Science , Beijing University of Posts and Telecommunications"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science , Beijing University of Posts and Telecommunications","institution_ids":["https://openalex.org/I139759216"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139720486","display_name":"Cong Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I168879160","display_name":"Zhejiang University of Science and Technology","ror":"https://ror.org/05mx0wr29","country_code":"CN","type":"education","lineage":["https://openalex.org/I168879160"]},{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Cong Wang","raw_affiliation_strings":["School of Control Science and Engineering , Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Control Science and Engineering , Zhejiang University","institution_ids":["https://openalex.org/I168879160","https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139730956","display_name":"Lidan Shou","orcid":null},"institutions":[{"id":"https://openalex.org/I4210156488","display_name":"Institute of Zoology of the Slovak Academy of Sciences","ror":"https://ror.org/05tm6j853","country_code":"SK","type":"facility","lineage":["https://openalex.org/I207624831","https://openalex.org/I4210156488"]},{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN","SK"],"is_corresponding":false,"raw_author_name":"Lidan Shou","raw_affiliation_strings":["Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","The State Key Laboratory of Blockchain and Data Security , Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","institution_ids":["https://openalex.org/I4210156488"]},{"raw_affiliation_string":"The State Key Laboratory of Blockchain and Data Security , Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139848747","display_name":"Gang Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Gang Chen","raw_affiliation_strings":["The State Key Laboratory of Blockchain and Data Security , Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The State Key Laboratory of Blockchain and Data Security , Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5139756701","display_name":"Huan Li","orcid":null},"institutions":[{"id":"https://openalex.org/I4210156488","display_name":"Institute of Zoology of the Slovak Academy of Sciences","ror":"https://ror.org/05tm6j853","country_code":"SK","type":"facility","lineage":["https://openalex.org/I207624831","https://openalex.org/I4210156488"]},{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN","SK"],"is_corresponding":false,"raw_author_name":"Huan Li","raw_affiliation_strings":["Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","The State Key Laboratory of Blockchain and Data Security , Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hangzhou High-Tech Zone (Binjiang ) Institute of Blockchain and Data Security","institution_ids":["https://openalex.org/I4210156488"]},{"raw_affiliation_string":"The State Key Laboratory of Blockchain and Data Security , Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.85428343,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"23707","last_page":"23726"},"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":0.3393999934196472,"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":0.3393999934196472,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.30469998717308044,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.08500000089406967,"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/decoding-methods","display_name":"Decoding methods","score":0.5827000141143799},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5519000291824341},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.35659998655319214},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.30160000920295715}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6621000170707703},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.5827000141143799},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5519000291824341},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47850000858306885},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.35659998655319214},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.34610000252723694},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3199999928474426},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.30160000920295715},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2831000089645386},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.2475000023841858}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.1087","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1087","pdf_url":"https://aclanthology.org/2026.acl-long.1087.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.1087","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.1087","pdf_url":"https://aclanthology.org/2026.acl-long.1087.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320318547","display_name":"Baidu","ror":"https://ror.org/03vs3wt56"},{"id":"https://openalex.org/F4320338464","display_name":"Natural Science Foundation of Zhejiang Province","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166880001.pdf","grobid_xml":"https://content.openalex.org/works/W7166880001.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Video":[0],"Large":[1],"Language":[2],"Models":[3],"(Video-LLMs)":[4],"excel":[5],"in":[6,56],"video":[7],"understanding":[8],"but":[9,105],"suffer":[10],"from":[11],"high":[12,113],"inference":[13],"latency":[14],"during":[15],"autoregressive":[16],"generation.Speculative":[17],"Decoding":[18],"(SD)":[19],"mitigates":[20],"this":[21,42,94],"by":[22,32,63,126,130,141],"applying":[23],"a":[24,78,96],"draft-and-verify":[25],"paradigm,":[26],"yet":[27],"existing":[28],"methods":[29,150],"are":[30],"constrained":[31],"rigid":[33],"exact-match":[34],"rules,":[35],"severely":[36],"limiting":[37],"the":[38,47,57,87,134],"acceleration":[39],"potential.To":[40],"bridge":[41],"gap,":[43],"we":[44,92],"propose":[45],"LVSPEC,":[46],"first":[48],"training-free":[49,148],"loosely":[50],"SD":[51,149],"framework":[52],"tailored":[53],"for":[54,151],"Video-LLMs.Grounded":[55],"insight":[58],"that":[59,100,110],"generation":[60],"is":[61,153],"governed":[62],"sparse":[64],"visual-relevant":[65,80],"anchors":[66],"(mandating":[67],"strictness)":[68],"amidst":[69],"abundant":[70],"visualirrelevant":[71],"fillers":[72],"(permitting":[73],"loose":[74],"verification),":[75],"LVSPEC":[76,111],"employs":[77],"lightweight":[79],"token":[81],"identification":[82],"scheme":[83],"to":[84,146],"accurately":[85],"pinpoint":[86],"former.To":[88],"further":[89],"maximize":[90],"acceptance,":[91],"augment":[93],"with":[95],"position-shift":[97],"tolerant":[98],"mechanism":[99],"effectively":[101],"salvages":[102],"positionally":[103],"mismatched":[104],"semantically":[106],"equivalent":[107],"tokens.Experiments":[108],"demonstrate":[109],"achieves":[112],"fidelity":[114],"and":[115,128,138,143],"speed:":[116],"it":[117,132],"preserves":[118],">99.8%":[119],"of":[120],"target":[121],"performance":[122],"while":[123],"accelerating":[124],"Qwen2.5-VL-32B":[125],"2.70\u00d7":[127],"LLaVA-OneVision-72B":[129],"2.94\u00d7.Notably,":[131],"boosts":[133],"mean":[135],"accepted":[136],"length":[137],"speedup":[139],"ratio":[140],"136%":[142],"35%":[144],"compared":[145],"SOTA":[147],"Video-LLMs.Code":[152],"at":[154],"https://github.com/zju-jiyicheng/LVSpec.":[155]},"counts_by_year":[],"updated_date":"2026-08-24T07:32:12.397491","created_date":"2026-07-02T00:00:00"}