{"id":"https://openalex.org/W7131405673","doi":"https://doi.org/10.48550/arxiv.2602.20566","title":"BFA++: Hierarchical Best-Feature-Aware Token Prune for Multi-View Vision Language Action Model","display_name":"BFA++: Hierarchical Best-Feature-Aware Token Prune for Multi-View Vision Language Action Model","publication_year":2026,"publication_date":"2026-02-24","ids":{"openalex":"https://openalex.org/W7131405673","doi":"https://doi.org/10.48550/arxiv.2602.20566"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.20566","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Li, Haosheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Li, Haosheng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005056360","display_name":"Weixin Mao","orcid":"https://orcid.org/0000-0002-0444-1079"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mao, Weixin","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126828360","display_name":"Zihan Lan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lan, Zihan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126808041","display_name":"Hongwei Xiong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiong, Hongwei","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126803743","display_name":"Hongan Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Hongan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124954044","display_name":"Chenyang Si","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Si, Chenyang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126806717","display_name":"Ziwei Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Ziwei","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102726348","display_name":"Xiaoming Deng","orcid":"https://orcid.org/0000-0003-4238-8288"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Deng, Xiaoming","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5126808533","display_name":"Hua Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Hua","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"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":0.7954000234603882,"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":0.7954000234603882,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.04769999906420708,"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/T10653","display_name":"Robot Manipulation and Learning","score":0.03590000048279762,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.7907000184059143},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.708299994468689},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.6000999808311462},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5593000054359436},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.4934999942779541},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.454800009727478},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.38679999113082886}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8224999904632568},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7907000184059143},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.708299994468689},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6133999824523926},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.6000999808311462},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5593000054359436},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.4934999942779541},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4684000015258789},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.454800009727478},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.38679999113082886},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.38519999384880066},{"id":"https://openalex.org/C144986985","wikidata":"https://www.wikidata.org/wiki/Q871236","display_name":"Hierarchical database model","level":2,"score":0.3490000069141388},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.31040000915527344},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2992999851703644},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.2973000109195709},{"id":"https://openalex.org/C2983448237","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Language understanding","level":2,"score":0.29339998960494995},{"id":"https://openalex.org/C5339829","wikidata":"https://www.wikidata.org/wiki/Q1425977","display_name":"Machine vision","level":2,"score":0.29330000281333923},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.2874000072479248},{"id":"https://openalex.org/C166109690","wikidata":"https://www.wikidata.org/wiki/Q4677422","display_name":"Action selection","level":3,"score":0.27950000762939453},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.25029999017715454}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.20566","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2602.20566","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.20566","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:doi:10.48550/arxiv.2602.20566","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"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":{"Vision-Language-Action":[0],"(VLA)":[1],"models":[2],"have":[3],"achieved":[4],"significant":[5],"breakthroughs":[6],"by":[7,103,178],"leveraging":[8],"Large":[9],"Vision":[10],"Language":[11],"Models":[12],"(VLMs)":[13],"to":[14,35,56,69,116,132],"jointly":[15],"interpret":[16],"instructions":[17],"and":[18,67,74,152,162,185,192,200,217],"visual":[19,26,145,212],"inputs.":[20],"However,":[21],"the":[22,62,72,159,175,183],"substantial":[23],"increase":[24],"in":[25,50,148,221],"tokens,":[27],"particularly":[28],"from":[29],"multi-view":[30],"inputs,":[31],"poses":[32],"serious":[33],"challenges":[34],"real-time":[36],"robotic":[37,78,164,223],"manipulation.":[38],"Existing":[39],"acceleration":[40],"techniques":[41],"for":[42,71,93],"VLMs,":[43],"such":[44],"as":[45,59,205],"token":[46,88,140,202],"pruning,":[47],"often":[48],"result":[49],"degraded":[51],"performance":[52],"when":[53],"directly":[54],"applied":[55],"VLA":[57,94],"models,":[58,187],"they":[60],"overlook":[61],"relationships":[63],"between":[64],"different":[65,129],"views":[66,127],"fail":[68],"account":[70],"dynamic":[73,87],"task-specific":[75],"characteristics":[76],"of":[77,190],"operation.":[79],"To":[80],"address":[81],"this,":[82],"we":[83],"propose":[84],"BFA++,":[85],"a":[86,98,206],"pruning":[89,100,203],"framework":[90],"designed":[91],"specifically":[92],"models.":[95],"BFA++":[96,168,173],"introduces":[97],"hierarchical":[99],"strategy":[101,209],"guided":[102],"two-level":[104],"importance":[105],"predictors:":[106],"an":[107,121],"intra-view":[108],"predictor":[109,123],"highlights":[110],"task-relevant":[111],"regions":[112],"within":[113],"each":[114],"image":[115],"suppress":[117],"spatial":[118],"noise,":[119],"while":[120,142],"inter-view":[122],"identifies":[124],"critical":[125],"camera":[126],"throughout":[128],"manipulation":[130,154,219],"phases":[131],"reduce":[133],"cross-view":[134],"redundancy.":[135],"This":[136],"design":[137],"enables":[138],"efficient":[139],"selection":[141],"preserving":[143],"essential":[144],"cues,":[146],"resulting":[147],"improved":[149,218],"computational":[150],"efficiency":[151],"higher":[153],"success":[155,176],"rates.":[156],"Evaluations":[157],"on":[158,181],"RoboTwin":[160],"benchmark":[161],"real-world":[163,222],"tasks":[165],"demonstrate":[166],"that":[167,198],"consistently":[169],"outperforms":[170],"existing":[171],"methods.":[172],"improves":[174],"rate":[177],"about":[179],"10%":[180],"both":[182],"\u03c00":[184],"RDT":[186],"achieving":[188],"speedup":[189],"1.8X":[191],"1.5X,":[193],"respectively.":[194],"Our":[195],"results":[196],"highlight":[197],"context-sensitive":[199],"task-aware":[201],"serves":[204],"more":[207],"effective":[208],"than":[210],"full":[211],"processing,":[213],"enabling":[214],"faster":[215],"inference":[216],"accuracy":[220],"systems.":[224]},"counts_by_year":[],"updated_date":"2026-04-04T16:13:02.066488","created_date":"2026-02-26T00:00:00"}