{"id":"https://openalex.org/W7160941174","doi":"https://doi.org/10.48550/arxiv.2605.08371","title":"PaceVGGT: Pre-Alternating-Attention Token Pruning for Visual Geometry Transformers","display_name":"PaceVGGT: Pre-Alternating-Attention Token Pruning for Visual Geometry Transformers","publication_year":2026,"publication_date":"2026-05-08","ids":{"openalex":"https://openalex.org/W7160941174","doi":"https://doi.org/10.48550/arxiv.2605.08371"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.08371","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08371","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.2605.08371","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135976188","display_name":"Haotang Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Haotang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136001258","display_name":"Zhenyu Qi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qi, Zhenyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135962070","display_name":"Shaohan Henry Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Shaohan Henry","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073436039","display_name":"Kebin Peng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Peng, Kebin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135986107","display_name":"Dr Zi Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135959150","display_name":"Qing Guo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Qing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135987775","display_name":"Sen He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Sen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5076154259","display_name":"Huanrui Yang","orcid":"https://orcid.org/0000-0002-3384-4512"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Huanrui","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/T10481","display_name":"Computer Graphics and Visualization Techniques","score":0.26440000534057617,"subfield":{"id":"https://openalex.org/subfields/1704","display_name":"Computer Graphics and Computer-Aided Design"},"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/T10481","display_name":"Computer Graphics and Visualization Techniques","score":0.26440000534057617,"subfield":{"id":"https://openalex.org/subfields/1704","display_name":"Computer Graphics and Computer-Aided Design"},"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":0.2409999966621399,"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/T10719","display_name":"3D Shape Modeling and Analysis","score":0.1485999971628189,"subfield":{"id":"https://openalex.org/subfields/2206","display_name":"Computational Mechanics"},"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":0.8058000206947327},{"id":"https://openalex.org/keywords/merge","display_name":"Merge (version control)","score":0.5878000259399414},{"id":"https://openalex.org/keywords/grid","display_name":"Grid","score":0.5310999751091003},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.47620001435279846},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.4546999931335449},{"id":"https://openalex.org/keywords/lexicographical-order","display_name":"Lexicographical order","score":0.39629998803138733},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.31139999628067017}],"concepts":[{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.8058000206947327},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6516000032424927},{"id":"https://openalex.org/C197129107","wikidata":"https://www.wikidata.org/wiki/Q1921621","display_name":"Merge (version control)","level":2,"score":0.5878000259399414},{"id":"https://openalex.org/C187691185","wikidata":"https://www.wikidata.org/wiki/Q2020720","display_name":"Grid","level":2,"score":0.5310999751091003},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.47620001435279846},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4546999931335449},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4512999951839447},{"id":"https://openalex.org/C159254197","wikidata":"https://www.wikidata.org/wiki/Q1144915","display_name":"Lexicographical order","level":2,"score":0.39629998803138733},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.32359999418258667},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.31139999628067017},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3091000020503998},{"id":"https://openalex.org/C175291020","wikidata":"https://www.wikidata.org/wiki/Q1156822","display_name":"Offset (computer science)","level":2,"score":0.2858999967575073},{"id":"https://openalex.org/C134757568","wikidata":"https://www.wikidata.org/wiki/Q274089","display_name":"Heap (data structure)","level":2,"score":0.28540000319480896},{"id":"https://openalex.org/C2780186347","wikidata":"https://www.wikidata.org/wiki/Q11414","display_name":"Subnetwork","level":2,"score":0.2842000126838684},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.265500009059906},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.2653000056743622},{"id":"https://openalex.org/C121684516","wikidata":"https://www.wikidata.org/wiki/Q7600677","display_name":"Computer graphics (images)","level":1,"score":0.26330000162124634},{"id":"https://openalex.org/C66882249","wikidata":"https://www.wikidata.org/wiki/Q169336","display_name":"Homogeneous","level":2,"score":0.25459998846054077},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.25369998812675476},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.2533000111579895}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.08371","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08371","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.2605.08371","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08371","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":[{"score":0.45623359084129333,"id":"https://metadata.un.org/sdg/11","display_name":"Sustainable cities and communities"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Visual":[0],"Geometry":[1],"Transformer":[2],"(VGGT)":[3],"is":[4,80],"a":[5,46,62,67,122,180],"strong":[6],"feed-forward":[7],"model":[8],"for":[9,184],"multiple":[10],"3D":[11],"tasks,":[12],"but":[13],"its":[14],"Alternating-Attention":[15],"(AA)":[16],"stack":[17],"scales":[18],"quadratically":[19],"in":[20],"the":[21,36,57,89,106,132,138,148],"total":[22,124],"token":[23,48],"count,":[24],"making":[25],"long":[26],"clips":[27],"expensive.":[28],"Existing":[29],"token-reduction":[30],"accelerators":[31],"operate":[32],"inside":[33],"AA,":[34],"leaving":[35],"patch":[37,54],"grid":[38,135],"that":[39,51,71],"enters":[40],"AA":[41,59],"uncompressed.":[42],"We":[43],"introduce":[44],"PaceVGGT,":[45],"pre-AA":[47,177],"pruning":[49,178],"framework":[50],"prunes":[52],"DINO":[53,76],"tokens":[55],"before":[56],"first":[58,81],"block":[60],"of":[61],"frozen":[63,185],"VGGT.":[64],"PaceVGGT":[65,145],"trains":[66],"lightweight":[68],"Token":[69],"Scorer":[70],"estimates":[72],"per-token":[73],"importance":[74],"from":[75,88,118],"features.":[77],"The":[78],"scorer":[79],"distilled":[82],"against":[83],"an":[84,111],"AA-internal":[85],"attention":[86],"target":[87],"unpruned":[90],"backbone,":[91],"then":[92],"refined":[93],"under":[94,121],"downstream":[95],"camera,":[96],"depth,":[97],"and":[98,143,168],"point-map":[99],"losses.":[100],"A":[101,127],"per-frame":[102],"keep":[103],"budget":[104],"fixes":[105],"backbone-visible":[107],"sequence":[108],"length,":[109],"while":[110,152],"importance-adaptive":[112],"merge/prune":[113],"assignment":[114],"preserves":[115],"residual":[116],"content":[117],"high-saliency":[119],"frames":[120],"fixed":[123],"merge":[125],"budget.":[126],"Feature-guided":[128],"Restoration":[129],"module":[130],"reconstructs":[131],"dense":[133],"spatial":[134],"required":[136],"by":[137,161],"prediction":[139],"heads.":[140],"On":[141,156],"ScanNet-50":[142],"7-Scenes,":[144],"remains":[146],"on":[147],"reconstruction":[149],"quality--latency":[150],"frontier":[151],"reducing":[153],"inference":[154],"latency.":[155],"ScanNet-50,":[157],"it":[158],"reduces":[159],"latency":[160],"\\(5.1\\times\\)":[162],"over":[163,170],"unmodified":[164],"VGGT":[165],"at":[166,172],"\\(N=300\\)":[167],"\\(1.47\\times\\)":[169],"LiteVGGT":[171],"\\(N=1000\\).":[173],"These":[174],"results":[175],"identify":[176],"as":[179],"viable":[181],"acceleration":[182],"route":[183],"VGGT-style":[186],"geometry":[187],"transformers.":[188]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-13T00:00:00"}