{"id":"https://openalex.org/W7155192339","doi":"https://doi.org/10.48550/arxiv.2604.19145","title":"ST-Prune: Training-Free Spatio-Temporal Token Pruning for Vision-Language Models in Autonomous Driving","display_name":"ST-Prune: Training-Free Spatio-Temporal Token Pruning for Vision-Language Models in Autonomous Driving","publication_year":2026,"publication_date":"2026-04-21","ids":{"openalex":"https://openalex.org/W7155192339","doi":"https://doi.org/10.48550/arxiv.2604.19145"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.19145","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.19145","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":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.2604.19145","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134284316","display_name":"Lin Sha","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sha, Lin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134290842","display_name":"Haiyun Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Haiyun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101956025","display_name":"Tao Wang","orcid":"https://orcid.org/0000-0002-3013-1947"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Tao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134352578","display_name":"Cong Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Cong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134238172","display_name":"Min Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Min","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134270237","display_name":"Jinqiao Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Jinqiao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134262710","display_name":"Qinghai Miao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Miao, Qinghai","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":0.7321000099182129,"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":0.7321000099182129,"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":0.10170000046491623,"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/T10531","display_name":"Advanced Vision and Imaging","score":0.047200001776218414,"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/security-token","display_name":"Security token","score":0.7075999975204468},{"id":"https://openalex.org/keywords/redundancy","display_name":"Redundancy (engineering)","score":0.7002999782562256},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.6527000069618225},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.47870001196861267},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4205999970436096},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.3587000072002411},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.3497999906539917}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7789999842643738},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7075999975204468},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.7002999782562256},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.6527000069618225},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5566999912261963},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.47870001196861267},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.42399999499320984},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4205999970436096},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.3587000072002411},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3497999906539917},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.3325999975204468},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.3109999895095825},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.31060001254081726},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.30709999799728394},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.28769999742507935},{"id":"https://openalex.org/C56951928","wikidata":"https://www.wikidata.org/wiki/Q3539213","display_name":"Trimming","level":2,"score":0.28200000524520874},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2752000093460083},{"id":"https://openalex.org/C154504017","wikidata":"https://www.wikidata.org/wiki/Q853614","display_name":"Identifier","level":2,"score":0.258899986743927},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.2524999976158142}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.19145","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.19145","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.19145","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.19145","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":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision-Language":[0],"Models":[1],"(VLMs)":[2],"have":[3],"become":[4],"central":[5],"to":[6,47,120,193],"autonomous":[7],"driving":[8,54],"systems,":[9],"yet":[10],"their":[11],"deployment":[12],"is":[13],"severely":[14],"bottlenecked":[15],"by":[16,84,114],"the":[17,49,95,116,185],"massive":[18],"computational":[19],"overhead":[20],"of":[21],"multi-view":[22],"camera":[23,118],"and":[24,44,75,88,102,128,161],"multi-frame":[25],"video":[26],"input.":[27],"Existing":[28],"token":[29,169,175],"pruning":[30,133,145,195],"methods,":[31],"primarily":[32],"designed":[33],"for":[34,167],"single-image":[35],"inputs,":[36],"treat":[37],"each":[38],"frame":[39],"or":[40],"view":[41],"in":[42,53],"isolation":[43],"thus":[45],"fail":[46],"exploit":[48],"inherent":[50],"spatio-temporal":[51,144],"redundancies":[52],"scenarios.":[55],"To":[56],"bridge":[57],"this":[58],"gap,":[59],"we":[60],"propose":[61],"ST-Prune,":[62],"a":[63,142],"training-free,":[64],"plug-and-play":[65],"framework":[66],"comprising":[67],"two":[68,138],"complementary":[69],"modules:":[70],"Motion-aware":[71],"Temporal":[72],"Pruning":[73,78],"(MTP)":[74],"Ring-view":[76],"Spatial":[77],"(RSP).":[79],"MTP":[80],"addresses":[81],"temporal":[82,89,132],"redundancy":[83,113],"encoding":[85],"motion":[86],"volatility":[87],"recency":[90],"as":[91],"soft":[92],"constraints":[93],"within":[94],"diversity":[96],"selection":[97],"objective,":[98],"prioritizing":[99],"dynamic":[100],"trajectories":[101],"current-frame":[103],"content":[104],"over":[105],"static":[106],"historical":[107],"background.":[108],"RSP":[109],"further":[110],"resolves":[111],"spatial":[112],"exploiting":[115],"ring-view":[117],"geometry":[119],"penalize":[121],"bilateral":[122],"cross-view":[123],"similarity,":[124],"eliminating":[125],"duplicate":[126],"projections":[127],"residual":[129],"background":[130],"that":[131],"alone":[134],"cannot":[135],"suppress.":[136],"These":[137],"modules":[139],"together":[140],"constitute":[141],"complete":[143],"process,":[146],"preserving":[147],"key":[148],"scene":[149],"information":[150],"under":[151],"strict":[152],"compression.":[153],"Validated":[154],"across":[155],"four":[156],"benchmarks":[157],"spanning":[158],"perception,":[159],"prediction,":[160],"planning,":[162],"ST-Prune":[163,177],"establishes":[164],"new":[165],"state-of-the-art":[166],"training-free":[168],"pruning.":[170],"Notably,":[171],"even":[172],"at":[173],"90\\%":[174],"reduction,":[176],"achieves":[178],"near-lossless":[179],"performance":[180],"with":[181],"certain":[182],"metrics":[183],"surpassing":[184],"full-model":[186],"baseline,":[187],"while":[188],"maintaining":[189],"inference":[190],"speeds":[191],"comparable":[192],"existing":[194],"approaches.":[196]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-04-23T00:00:00"}