{"id":"https://openalex.org/W4403564591","doi":"https://doi.org/10.48550/arxiv.2410.09324","title":"Token Pruning using a Lightweight Background Aware Vision Transformer","display_name":"Token Pruning using a Lightweight Background Aware Vision Transformer","publication_year":2024,"publication_date":"2024-10-12","ids":{"openalex":"https://openalex.org/W4403564591","doi":"https://doi.org/10.48550/arxiv.2410.09324"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2410.09324","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.09324","pdf_url":"https://arxiv.org/pdf/2410.09324","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2410.09324","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5109935279","display_name":"Sudhakar Sah","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sah, Sudhakar","raw_affiliation_strings":["Toronto , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Toronto , Canada","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103734348","display_name":"Ravish Kumar","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kumar, Ravish","raw_affiliation_strings":["Toronto , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Toronto , Canada","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024979989","display_name":"Honnesh Rohmetra","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rohmetra, Honnesh","raw_affiliation_strings":["Toronto , Canada"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Toronto , Canada","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5112441744","display_name":"Ehsan Saboori","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Saboori, Ehsan","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":true,"cited_by_count":1,"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":1.0,"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":1.0,"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/T11605","display_name":"Visual Attention and Saliency Detection","score":0.0,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/token-pruning","display_name":"token pruning","score":0.9990000128746033},{"id":"https://openalex.org/keywords/detr","display_name":"DETR","score":0.8040000200271606},{"id":"https://openalex.org/keywords/edge-intelligence","display_name":"edge AI","score":0.5600000023841858},{"id":"https://openalex.org/keywords/inference-throughput","display_name":"inference throughput","score":0.5600000023841858},{"id":"https://openalex.org/keywords/memory-reduction","display_name":"memory reduction","score":0.6869999766349792},{"id":"https://openalex.org/keywords/segmentation-maps","display_name":"segmentation maps","score":0.33799999952316284},{"id":"https://openalex.org/keywords/bounding-box-annotation","display_name":"bounding box annotation","score":0.7929999828338623}],"concepts":[{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7362909913063049},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.6239702701568604},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5524340867996216},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.4534643888473511},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32694071531295776},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.16512268781661987},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.16173279285430908},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.09511059522628784},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.06372159719467163},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.05557623505592346},{"id":"https://openalex.org/C6557445","wikidata":"https://www.wikidata.org/wiki/Q173113","display_name":"Agronomy","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2410.09324","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.09324","pdf_url":"https://arxiv.org/pdf/2410.09324","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2410.09324","doi":"https://doi.org/10.48550/arxiv.2410.09324","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2410.09324","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"pmh:oai:arXiv.org:2410.09324","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.09324","pdf_url":"https://arxiv.org/pdf/2410.09324","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4403564591.pdf","grobid_xml":"https://content.openalex.org/works/W4403564591.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4388335561","https://openalex.org/W2970530566","https://openalex.org/W4288261899","https://openalex.org/W4307309205","https://openalex.org/W2967478618","https://openalex.org/W4385009901","https://openalex.org/W4304700937"],"abstract_inverted_index":{"High":[0],"runtime":[1,58],"memory":[2,59],"and":[3,13,60,123,129,137,143,145,181],"high":[4],"latency":[5],"puts":[6],"significant":[7],"constraint":[8],"on":[9,16,31,140,148],"Vision":[10,42],"Transformer":[11,43],"training":[12],"inference,":[14],"especially":[15],"edge":[17],"devices.":[18],"Token":[19],"pruning":[20],"reduces":[21],"the":[22,28,73,94,165],"number":[23],"of":[24,34,110,126,175],"input":[25],"tokens":[26,71,77,114,131],"to":[27,49,56,68,86,105,112,115,161],"ViT":[29,88,111],"based":[30,89],"importance":[32],"criteria":[33],"each":[35],"token.":[36],"We":[37,92,152],"present":[38],"a":[39,46,65,87,107,154,172],"Background":[40],"Aware":[41],"(BAViT)":[44],"model,":[45],"pre-processing":[47],"block":[48],"object":[50,90],"detection":[51],"models":[52],"like":[53],"DETR/YOLOS":[54],"aimed":[55],"reduce":[57],"increase":[61,164],"throughput":[62,166],"by":[63,98,167],"using":[64],"novel":[66],"approach":[67,187],"identify":[69],"background":[70,76,128],"in":[72],"image.":[74],"The":[75],"can":[78,132,163],"be":[79,133],"pruned":[80],"completely":[81],"or":[82,118],"partially":[83],"before":[84],"feeding":[85],"detector.":[91],"use":[93,194],"semantic":[95],"information":[96],"provided":[97],"segmentation":[99],"map":[100],"and/or":[101],"bounding":[102],"box":[103],"annotation":[104],"train":[106],"few":[108],"layers":[109,122,125],"classify":[113],"either":[116],"foreground":[117,130],"background.":[119],"Using":[120],"2":[121,155],"10":[124],"BAViT,":[127],"separated":[134],"with":[135,171,183],"75%":[136],"88%":[138],"accuracy":[139,147],"VOC":[141],"dataset":[142,150],"71%":[144],"80%":[146],"COCO":[149],"respectively.":[151],"show":[153],"layer":[156],"BAViT-small":[157],"model":[158],"as":[159],"pre-processor":[160],"YOLOS":[162],"30%":[168],"-":[169],"40%":[170],"mAP":[173],"drop":[174],"3%":[176],"without":[177],"any":[178],"sparse":[179,184],"fine-tuning":[180],"2%":[182],"fine-tuning.":[185],"Our":[186],"is":[188],"specifically":[189],"targeted":[190],"for":[191],"Edge":[192],"AI":[193],"cases.":[195]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2025-10-10T00:00:00"}