{"id":"https://openalex.org/W7154698039","doi":"https://doi.org/10.48550/arxiv.2604.14563","title":"Revisiting Token Compression for Accelerating ViT-based Sparse Multi-View 3D Object Detectors","display_name":"Revisiting Token Compression for Accelerating ViT-based Sparse Multi-View 3D Object Detectors","publication_year":2026,"publication_date":"2026-04-16","ids":{"openalex":"https://openalex.org/W7154698039","doi":"https://doi.org/10.48550/arxiv.2604.14563"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.14563","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.14563","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.2604.14563","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101348403","display_name":"Mingqian Ji","orcid":"https://orcid.org/0009-0006-1855-1945"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji, Mingqian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133837523","display_name":"Shanshan Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Shanshan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133889998","display_name":"Jian Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Jian","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/T10036","display_name":"Advanced Neural Network Applications","score":0.9232000112533569,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9232000112533569,"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/T10812","display_name":"Human Pose and Action Recognition","score":0.008899999782443047,"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/T11448","display_name":"Face recognition and analysis","score":0.006399999838322401,"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":0.7008000016212463},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6316999793052673},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.47040000557899475},{"id":"https://openalex.org/keywords/feature-selection","display_name":"Feature selection","score":0.44859999418258667},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.44339999556541443},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.4433000087738037},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.439300000667572},{"id":"https://openalex.org/keywords/object-detection","display_name":"Object detection","score":0.4348999857902527},{"id":"https://openalex.org/keywords/detector","display_name":"Detector","score":0.4343000054359436}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7925000190734863},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7008000016212463},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6316999793052673},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5091000199317932},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.47040000557899475},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.44859999418258667},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.44339999556541443},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4433000087738037},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.439300000667572},{"id":"https://openalex.org/C2776151529","wikidata":"https://www.wikidata.org/wiki/Q3045304","display_name":"Object detection","level":3,"score":0.4348999857902527},{"id":"https://openalex.org/C94915269","wikidata":"https://www.wikidata.org/wiki/Q1834857","display_name":"Detector","level":2,"score":0.4343000054359436},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.39480000734329224},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3799000084400177},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.3781999945640564},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3693000078201294},{"id":"https://openalex.org/C25797200","wikidata":"https://www.wikidata.org/wiki/Q828137","display_name":"Compression ratio","level":3,"score":0.36899998784065247},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.3621000051498413},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.36149999499320984},{"id":"https://openalex.org/C2781238097","wikidata":"https://www.wikidata.org/wiki/Q175026","display_name":"Object (grammar)","level":2,"score":0.3610999882221222},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.36059999465942383},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.335099995136261},{"id":"https://openalex.org/C180016635","wikidata":"https://www.wikidata.org/wiki/Q2712821","display_name":"Compression (physics)","level":2,"score":0.3109999895095825},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.29409998655319214},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.29030001163482666},{"id":"https://openalex.org/C2781122975","wikidata":"https://www.wikidata.org/wiki/Q16928266","display_name":"Semantic feature","level":2,"score":0.2888999879360199},{"id":"https://openalex.org/C64876066","wikidata":"https://www.wikidata.org/wiki/Q5141226","display_name":"Cognitive neuroscience of visual object recognition","level":3,"score":0.2750999927520752},{"id":"https://openalex.org/C115051666","wikidata":"https://www.wikidata.org/wiki/Q6522493","display_name":"Ranging","level":2,"score":0.26919999718666077},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.26649999618530273}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.14563","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.14563","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.2604.14563","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.14563","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":{"Vision":[0],"Transformer":[1],"(ViT)-based":[2],"sparse":[3],"multi-view":[4],"3D":[5,66],"object":[6],"detectors":[7],"have":[8],"achieved":[9],"remarkable":[10],"accuracy":[11],"but":[12],"still":[13],"suffer":[14],"from":[15],"high":[16],"inference":[17,171],"latency":[18],"due":[19],"to":[20,103,108,115,118,168],"heavy":[21],"token":[22,28,42],"processing.":[23],"To":[24,68,122],"accelerate":[25],"these":[26,70],"models,":[27],"compression":[29],"has":[30],"been":[31],"widely":[32],"explored.":[33],"However,":[34],"our":[35],"revisit":[36],"of":[37],"existing":[38],"strategies,":[39],"such":[40],"as":[41],"pruning,":[43],"merging,":[44],"and":[45,60,112,139,158,176],"patch":[46,81],"size":[47],"enlargement,":[48],"reveals":[49],"that":[50,78,99,164],"they":[51],"often":[52],"discard":[53],"informative":[54,134],"background":[55],"cues,":[56],"disrupt":[57],"contextual":[58],"consistency,":[59],"lose":[61],"fine-grained":[62,145],"semantics,":[63],"negatively":[64],"affecting":[65],"detection.":[67],"overcome":[69],"limitations,":[71],"we":[72,92],"propose":[73],"SEPatch3D,":[74],"a":[75],"novel":[76],"framework":[77],"dynamically":[79],"adjusts":[80],"sizes":[82],"while":[83,184],"preserving":[84,185],"critical":[85],"semantic":[86,152],"information":[87],"within":[88],"coarse":[89,149],"patches.":[90],"Specifically,":[91],"design":[93],"Spatiotemporal-aware":[94],"Patch":[95,129],"Size":[96],"Selection":[97,130],"(SPSS)":[98],"assigns":[100],"small":[101],"patches":[102,114,135],"scenes":[104,117],"containing":[105],"nearby":[106],"objects":[107],"preserve":[109],"fine":[110],"details":[111,146],"large":[113],"background-dominated":[116],"reduce":[119],"computation":[120],"cost.":[121],"further":[123],"mitigate":[124],"potential":[125],"detail":[126],"loss,":[127],"Informative":[128],"(IPS)":[131],"selects":[132],"the":[133,156,173,181],"for":[136],"feature":[137],"refinement,":[138],"Cross-Granularity":[140],"Feature":[141],"Enhancement":[142],"(CGFE)":[143],"injects":[144],"into":[147],"selected":[148],"patches,":[150],"enriching":[151],"features.":[153],"Experiments":[154],"on":[155],"nuScenes":[157],"Argoverse":[159],"2":[160],"validation":[161],"sets":[162],"show":[163],"SEPatch3D":[165],"achieves":[166],"up":[167],"\\textbf{57\\%}":[169],"faster":[170],"than":[172,180],"StreamPETR":[174],"baseline":[175],"\\textbf{20\\%}":[177],"higher":[178],"efficiency":[179],"state-of-the-art":[182],"ToC3D-faster,":[183],"comparable":[186],"detection":[187],"accuracy.":[188],"Code":[189],"is":[190],"available":[191],"at":[192],"https://github.com/Mingqj/SEPatch3D.":[193]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-04-18T00:00:00"}