{"id":"https://openalex.org/W7161745907","doi":"https://doi.org/10.48550/arxiv.2605.18041","title":"OmniSelect: Dynamic Modality-Aware Token Compression for Efficient Omni-modal Large Language Models","display_name":"OmniSelect: Dynamic Modality-Aware Token Compression for Efficient Omni-modal Large Language Models","publication_year":2026,"publication_date":"2026-05-18","ids":{"openalex":"https://openalex.org/W7161745907","doi":"https://doi.org/10.48550/arxiv.2605.18041"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.18041","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18041","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":"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.18041","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135857399","display_name":"Morunliu Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Morunliu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037627177","display_name":"Ruotao Xu","orcid":"https://orcid.org/0000-0002-5277-9859"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Ruotao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136495388","display_name":"Le Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Le","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136488626","display_name":"Yue Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yue","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136459026","display_name":"Jianxin Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Jianxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136497658","display_name":"Juntao Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Juntao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063901345","display_name":"Yihang Lou","orcid":"https://orcid.org/0000-0002-8143-389X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lou, Yihang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136482243","display_name":"Siwei Feng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feng, Siwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136498741","display_name":"Peifeng Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Peifeng","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":0.42660000920295715,"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":0.42660000920295715,"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/T10860","display_name":"Speech and Audio Processing","score":0.24150000512599945,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T11309","display_name":"Music and Audio Processing","score":0.09120000153779984,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/security-token","display_name":"Security token","score":0.8583999872207642},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.7056000232696533},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.6190000176429749},{"id":"https://openalex.org/keywords/relevance","display_name":"Relevance (law)","score":0.6032000184059143},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.45890000462532043},{"id":"https://openalex.org/keywords/abstraction","display_name":"Abstraction","score":0.438400000333786},{"id":"https://openalex.org/keywords/reduction","display_name":"Reduction (mathematics)","score":0.41690000891685486}],"concepts":[{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.8583999872207642},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8198999762535095},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.7056000232696533},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.6190000176429749},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.6032000184059143},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5443999767303467},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.45890000462532043},{"id":"https://openalex.org/C124304363","wikidata":"https://www.wikidata.org/wiki/Q673661","display_name":"Abstraction","level":2,"score":0.438400000333786},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.41690000891685486},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.40369999408721924},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.382099986076355},{"id":"https://openalex.org/C115067241","wikidata":"https://www.wikidata.org/wiki/Q1639854","display_name":"Token passing","level":3,"score":0.37220001220703125},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.34950000047683716},{"id":"https://openalex.org/C2779903281","wikidata":"https://www.wikidata.org/wiki/Q6888026","display_name":"Modalities","level":2,"score":0.3273000121116638},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.3133000135421753},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.3127000033855438},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2736000120639801},{"id":"https://openalex.org/C2780226545","wikidata":"https://www.wikidata.org/wiki/Q6888030","display_name":"Modality (human\u2013computer interaction)","level":2,"score":0.25529998540878296},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.25429999828338623}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.18041","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18041","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.18041","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18041","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":"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Omnimodal":[0],"large":[1],"language":[2],"models":[3],"(OmniLLMs)":[4],"have":[5],"recently":[6],"gained":[7],"increasing":[8],"attention":[9],"for":[10,41,69],"unified":[11],"audio-video":[12],"understanding.":[13],"However,":[14],"processing":[15],"long":[16],"multimodal":[17,70,147],"token":[18,26,60,105,148],"sequences":[19],"introduces":[20],"substantial":[21],"computational":[22],"overhead,":[23],"making":[24],"efficient":[25,146],"compression":[27,67],"crucial.":[28],"Existing":[29],"methods":[30],"typically":[31],"rely":[32],"on":[33,97],"fixed,":[34],"modality-specific":[35],"guidance,":[36],"which":[37],"fails":[38],"to":[39,79,115],"account":[40],"the":[42,134],"varying":[43],"importance":[44],"of":[45,136],"modalities":[46],"across":[47,119],"different":[48],"queries.":[49],"To":[50],"address":[51],"this":[52],"limitation,":[53],"we":[54,73],"propose":[55],"$\\textbf{OmniSelect}$,":[56],"a":[57,75],"training-free,":[58],"modality-adaptive":[59],"pruning":[61,89,106,113],"framework":[62],"that":[63,142],"dynamically":[64],"selects":[65],"appropriate":[66],"strategies":[68],"inputs.":[71],"Specifically,":[72],"leverage":[74],"lightweight":[76],"AudioCLIP":[77],"model":[78],"estimate":[80],"cross-modal":[81],"relevance":[82,99],"and":[83,93,126],"categorize":[84],"each":[85,108],"input":[86],"into":[87],"three":[88],"regimes:":[90],"Audio-Centric,":[91],"Video-Centric,":[92],"Uniform":[94],"pruning.":[95],"Based":[96],"these":[98],"scores,":[100],"OmniSelect":[101,131],"further":[102],"performs":[103],"fine-grained":[104],"within":[107],"temporal":[109],"group,":[110],"adaptively":[111],"allocating":[112],"ratios":[114],"preserve":[116],"informative":[117],"tokens":[118],"modalities.":[120],"By":[121],"explicitly":[122],"modeling":[123],"modality":[124],"preference":[125],"enabling":[127],"dynamic":[128],"strategy":[129],"selection,":[130],"effectively":[132],"avoids":[133],"pitfalls":[135],"one-size-fits-all":[137],"compression.":[138],"Extensive":[139],"experiments":[140],"demonstrate":[141],"our":[143],"method":[144],"achieves":[145],"reduction":[149],"while":[150],"maintaining":[151],"strong":[152],"performance,":[153],"without":[154],"requiring":[155],"any":[156],"additional":[157],"training.":[158]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-20T00:00:00"}