{"id":"https://openalex.org/W6979316813","doi":"https://doi.org/10.1007/978-3-032-31666-0_43","title":"LinDeps: A Fine-tuning Free Post-Pruning Method to Remove Layer-Wise Linear Dependencies with Guaranteed Performance Preservation","display_name":"LinDeps: A Fine-tuning Free Post-Pruning Method to Remove Layer-Wise Linear Dependencies with Guaranteed Performance Preservation","publication_year":2025,"publication_date":"2025-07-29","ids":{"openalex":"https://openalex.org/W6979316813","doi":"https://doi.org/10.1007/978-3-032-31666-0_43"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2507.21573","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2507.21573","pdf_url":"https://arxiv.org/pdf/2507.21573","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,"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"type":"preprint","indexed_in":["arxiv"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2507.21573","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Henry, Maxim","orcid":null},"institutions":[{"id":"https://openalex.org/I157674565","display_name":"University of Li\u00e8ge","ror":"https://ror.org/00afp2z80","country_code":"BE","type":"education","lineage":["https://openalex.org/I157674565"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Henry, Maxim","raw_affiliation_strings":["Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium","institution_ids":["https://openalex.org/I157674565"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Deli\u00e8ge, Adrien","orcid":null},"institutions":[{"id":"https://openalex.org/I157674565","display_name":"University of Li\u00e8ge","ror":"https://ror.org/00afp2z80","country_code":"BE","type":"education","lineage":["https://openalex.org/I157674565"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Deli\u00e8ge, Adrien","raw_affiliation_strings":["Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium","institution_ids":["https://openalex.org/I157674565"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Cioppa, Anthony","orcid":null},"institutions":[{"id":"https://openalex.org/I157674565","display_name":"University of Li\u00e8ge","ror":"https://ror.org/00afp2z80","country_code":"BE","type":"education","lineage":["https://openalex.org/I157674565"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Cioppa, Anthony","raw_affiliation_strings":["Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium","institution_ids":["https://openalex.org/I157674565"]}]},{"author_position":"last","author":{"id":null,"display_name":"Van Droogenbroeck, Marc","orcid":null},"institutions":[{"id":"https://openalex.org/I157674565","display_name":"University of Li\u00e8ge","ror":"https://ror.org/00afp2z80","country_code":"BE","type":"education","lineage":["https://openalex.org/I157674565"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Van Droogenbroeck, Marc","raw_affiliation_strings":["Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Montefiore Institute , University of Li\u00e8ge , Li\u00e8ge , Belgium","institution_ids":["https://openalex.org/I157674565"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I157674565"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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":true,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.8892999887466431,"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":0.8892999887466431,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.01269999984651804,"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"}},{"id":"https://openalex.org/T11019","display_name":"Image Enhancement Techniques","score":0.011900000274181366,"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/pruning","display_name":"Pruning","score":0.8101000189781189},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.5733000040054321},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.48570001125335693},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.4733000099658966},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4618000090122223},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.44519999623298645},{"id":"https://openalex.org/keywords/computational-complexity-theory","display_name":"Computational complexity theory","score":0.4081999957561493}],"concepts":[{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.8101000189781189},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7253000140190125},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.5733000040054321},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.49959999322891235},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.48570001125335693},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.4733000099658966},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4618000090122223},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.44519999623298645},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.4081999957561493},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3882000148296356},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.352400004863739},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.33239999413490295},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.3125999867916107},{"id":"https://openalex.org/C2944601119","wikidata":"https://www.wikidata.org/wiki/Q43744058","display_name":"Residual neural network","level":3,"score":0.2822999954223633},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.27720001339912415},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.27459999918937683},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.25870001316070557}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:arXiv.org:2507.21573","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2507.21573","pdf_url":"https://arxiv.org/pdf/2507.21573","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,"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2507.21573","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2507.21573","pdf_url":"https://arxiv.org/pdf/2507.21573","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,"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321390","display_name":"Fonds De La Recherche Scientifique - FNRS","ror":"https://ror.org/03q83t159"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W6979316813.pdf","grobid_xml":"https://content.openalex.org/works/W6979316813.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Convolutional":[0],"Neural":[1],"Networks":[2],"(CNN)":[3],"are":[4],"widely":[5],"used":[6],"in":[7,59,80,187,194],"many":[8],"computer":[9],"vision":[10],"tasks.":[11],"Yet,":[12],"their":[13],"increasing":[14],"size":[15,39],"and":[16,40,111,130,149,159,163,208],"complexity":[17],"pose":[18],"significant":[19,205],"challenges":[20],"for":[21,221],"efficient":[22],"deployment":[23],"on":[24,102,157],"resource-constrained":[25],"platforms.":[26],"Hence,":[27],"network":[28],"pruning":[29,57,68,82,96,106,174,206,226],"has":[30],"emerged":[31],"as":[32],"an":[33,218],"effective":[34],"way":[35],"of":[36,43,104,172,184],"reducing":[37],"the":[38,142,185],"computational":[41],"requirements":[42],"neural":[44],"networks":[45],"by":[46],"removing":[47,61],"redundant":[48,113],"or":[49,224],"unimportant":[50],"parameters.":[51],"However,":[52],"a":[53,77,90,95,136,181,212],"fundamental":[54],"challenge":[55],"with":[56,161],"consists":[58],"optimally":[60],"redundancies":[62],"without":[63,151],"degrading":[64],"performance.":[65],"Most":[66],"existing":[67,173],"techniques":[69,175],"overlook":[70],"structural":[71],"dependencies":[72],"across":[73],"feature":[74,126],"maps":[75,127],"within":[76],"layer,":[78],"resulting":[79],"suboptimal":[81],"decisions.":[83],"In":[84],"this":[85],"work,":[86],"we":[87],"introduce":[88],"LinDeps,":[89],"novel":[91,137],"post-pruning":[92],"method,":[93],"i.e.,":[94],"method":[97],"that":[98,167],"can":[99,200],"be":[100,201],"applied":[101],"top":[103],"any":[105,153,222],"technique,":[107],"which":[108,203],"systematically":[109],"identifies":[110],"removes":[112],"filters":[114],"via":[115],"linear":[116],"dependency":[117],"analysis.":[118],"Particularly,":[119],"LinDeps":[120,168,193,215],"applies":[121],"pivoted":[122],"QR":[123],"decomposition":[124],"to":[125,128,146,180],"detect":[129],"prune":[131],"linearly":[132],"dependent":[133],"filters.":[134],"Then,":[135],"signal":[138],"recovery":[139],"mechanism":[140],"adjusts":[141],"next":[143],"layer's":[144],"kernels":[145],"preserve":[147],"compatibility":[148],"performance":[150],"requiring":[152],"fine-tuning.":[154],"Our":[155],"experiments":[156],"CIFAR-10":[158],"ImageNet":[160],"VGG":[162],"ResNet":[164],"backbones":[165],"demonstrate":[166],"improves":[169],"compression":[170],"rates":[171],"while":[176],"preserving":[177],"performances,":[178],"leading":[179],"new":[182],"state":[183],"art":[186],"CNN":[188],"pruning.":[189],"We":[190],"also":[191],"benchmark":[192],"low-resource":[195],"setups":[196],"where":[197],"no":[198],"retraining":[199],"performed,":[202],"shows":[204],"improvements":[207],"inference":[209],"speedups":[210],"over":[211],"state-of-the-art":[213],"method.":[214],"therefore":[216],"constitutes":[217],"essential":[219],"add-on":[220],"current":[223],"future":[225],"technique.":[227]},"counts_by_year":[],"updated_date":"2026-08-26T07:47:46.906454","created_date":"2025-10-10T00:00:00"}