{"id":"https://openalex.org/W4415966836","doi":"https://doi.org/10.1109/iccv51701.2025.02208","title":"SparseVILA: Decoupling Visual Sparsity for Efficient VLM Inference","display_name":"SparseVILA: Decoupling Visual Sparsity for Efficient VLM Inference","publication_year":2025,"publication_date":"2025-10-19","ids":{"openalex":"https://openalex.org/W4415966836","doi":"https://doi.org/10.1109/iccv51701.2025.02208"},"language":null,"primary_location":{"id":"doi:10.1109/iccv51701.2025.02208","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccv51701.2025.02208","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/CVF International Conference on Computer Vision (ICCV)","raw_type":"proceedings-article"},"type":"article","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2510.17777","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5036842123","display_name":"Samir Khaki","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Khaki, Samir","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088908535","display_name":"Junxian Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Junxian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103244882","display_name":"Jiaming Tang","orcid":"https://orcid.org/0009-0004-4186-6561"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Jiaming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112039102","display_name":"Shang Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Shang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5060855982","display_name":"Yukang Chen","orcid":"https://orcid.org/0000-0002-5933-2654"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Yukang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059152392","display_name":"Konstantinos N. Plataniotis","orcid":"https://orcid.org/0000-0003-3647-5473"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Plataniotis, Konstantinos N.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037770584","display_name":"Yao Lu","orcid":"https://orcid.org/0000-0002-1094-8562"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Yao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100623629","display_name":"Song Han","orcid":"https://orcid.org/0000-0002-3951-8005"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Song","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5065136413","display_name":"Zhijian Liu","orcid":"https://orcid.org/0009-0007-3905-9893"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zhijian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5036842123"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.30645113,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"23784","last_page":"23794"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9944000244140625,"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":0.9944000244140625,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.0012000000569969416,"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":0.0007999999797903001,"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/inference","display_name":"Inference","score":0.7297999858856201},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.6438000202178955},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.612500011920929},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5669000148773193},{"id":"https://openalex.org/keywords/conversation","display_name":"Conversation","score":0.45350000262260437},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.4041000008583069},{"id":"https://openalex.org/keywords/decoupling","display_name":"Decoupling (probability)","score":0.375},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.37059998512268066},{"id":"https://openalex.org/keywords/cache","display_name":"Cache","score":0.33660000562667847}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.816100001335144},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7297999858856201},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.6438000202178955},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.612500011920929},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6003000140190125},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5669000148773193},{"id":"https://openalex.org/C2777200299","wikidata":"https://www.wikidata.org/wiki/Q52943","display_name":"Conversation","level":2,"score":0.45350000262260437},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41040000319480896},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.4041000008583069},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.375},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.37059998512268066},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3402999937534332},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.33660000562667847},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.3149999976158142},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3050000071525574},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.30480000376701355},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.2838999927043915},{"id":"https://openalex.org/C2777472644","wikidata":"https://www.wikidata.org/wiki/Q16968992","display_name":"Approximate inference","level":3,"score":0.28209999203681946},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.28029999136924744},{"id":"https://openalex.org/C2780103172","wikidata":"https://www.wikidata.org/wiki/Q1309721","display_name":"Visual Objects","level":3,"score":0.2775000035762787},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2644999921321869},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2612999975681305},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.257999986410141},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2574999928474426},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.2567000091075897},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.2547000050544739},{"id":"https://openalex.org/C113364801","wikidata":"https://www.wikidata.org/wiki/Q26674","display_name":"High fidelity","level":2,"score":0.2538999915122986}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1109/iccv51701.2025.02208","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccv51701.2025.02208","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/CVF International Conference on Computer Vision (ICCV)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2510.17777","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.17777","pdf_url":"https://arxiv.org/pdf/2510.17777","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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.2510.17777","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.17777","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":"article"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2510.17777","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.17777","pdf_url":"https://arxiv.org/pdf/2510.17777","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4415966836.pdf"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision":[0],"Language":[1],"Models":[2],"(VLMs)":[3],"have":[4],"rapidly":[5],"advanced":[6],"in":[7],"integrating":[8],"visual":[9,34,52,68,96],"and":[10,21,57,72,127,144,151],"textual":[11],"reasoning,":[12],"powering":[13],"applications":[14],"across":[15,54,63],"high-resolution":[16],"image":[17],"understanding,":[18],"long-video":[19],"analysis,":[20],"multi-turn":[22,89],"conversation.":[23],"However,":[24],"their":[25],"scalability":[26],"remains":[27],"limited":[28],"by":[29,65,91],"the":[30,55,95],"growing":[31],"number":[32],"of":[33,94],"tokens":[35,69,76,101],"that":[36,50,99],"dominate":[37],"inference":[38,49,113],"latency.":[39],"We":[40],"present":[41],"SparseVILA,":[42],"a":[43,156,164],"new":[44,157],"paradigm":[45],"for":[46,159,168],"efficient":[47,160],"VLM":[48],"decouples":[51],"sparsity":[53,62],"prefilling":[56],"decoding":[58],"stages.":[59],"SparseVILA":[60,115,154],"distributes":[61],"stages":[64],"pruning":[66,85,150],"redundant":[67],"during":[70,77],"prefill":[71,84],"retrieving":[73],"only":[74],"query-relevant":[75],"decoding.":[78],"This":[79],"decoupled":[80],"design":[81],"matches":[82],"leading":[83],"methods":[86],"while":[87,139],"preserving":[88],"fidelity":[90],"retaining":[92],"most":[93],"cache":[97],"so":[98],"query-aware":[100,152],"can":[102],"be":[103],"retrieved":[104],"at":[105],"each":[106],"conversation":[107],"round.":[108],"Built":[109],"on":[110,134,142],"an":[111,128],"AWQ-optimized":[112],"pipeline,":[114],"achieves":[116],"up":[117],"to":[118],"4.0":[119],"times":[120,124,131],"faster":[121,125],"prefilling,":[122],"2.5":[123],"decoding,":[126],"overall":[129],"2.6":[130],"end-to-end":[132],"speedup":[133],"long-context":[135],"video":[136],"tasks":[137],"--":[138],"improving":[140],"accuracy":[141],"document-understanding":[143],"reasoning":[145],"tasks.":[146],"By":[147],"decoupling":[148],"query-agnostic":[149],"retrieval,":[153],"establishes":[155],"direction":[158],"multimodal":[161],"inference,":[162],"offering":[163],"training-free,":[165],"architecture-agnostic":[166],"framework":[167],"accelerating":[169],"large":[170],"VLMs":[171],"without":[172],"sacrificing":[173],"capability.":[174]},"counts_by_year":[],"updated_date":"2026-05-05T08:41:31.759640","created_date":"2025-10-22T00:00:00"}