{"id":"https://openalex.org/W7155091643","doi":"https://doi.org/10.48550/arxiv.2604.16834","title":"Towards Deep Encrypted Training: Low-Latency, Memory-Efficient, and High-Throughput Inference for Privacy-Preserving Neural Networks","display_name":"Towards Deep Encrypted Training: Low-Latency, Memory-Efficient, and High-Throughput Inference for Privacy-Preserving Neural Networks","publication_year":2026,"publication_date":"2026-04-18","ids":{"openalex":"https://openalex.org/W7155091643","doi":"https://doi.org/10.48550/arxiv.2604.16834"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.16834","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.16834","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.16834","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5114654076","display_name":"Nges Brian Njungle","orcid":"https://orcid.org/0009-0006-3393-6851"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Njungle, Nges Brian","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5022113521","display_name":"Eric Jahns","orcid":"https://orcid.org/0009-0004-5511-7975"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jahns, Eric","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5069200437","display_name":"Michel A. Kinsy","orcid":"https://orcid.org/0000-0002-1432-6939"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kinsy, Michel A.","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5114654076"],"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/T10237","display_name":"Cryptography and Data Security","score":0.5871999859809875,"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"}},"topics":[{"id":"https://openalex.org/T10237","display_name":"Cryptography and Data Security","score":0.5871999859809875,"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/T11017","display_name":"Chaos-based Image/Signal Encryption","score":0.26989999413490295,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.041200000792741776,"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/inference","display_name":"Inference","score":0.7297000288963318},{"id":"https://openalex.org/keywords/encryption","display_name":"Encryption","score":0.7131999731063843},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.6341999769210815},{"id":"https://openalex.org/keywords/homomorphic-encryption","display_name":"Homomorphic encryption","score":0.5807999968528748},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5146999955177307},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.4221000075340271},{"id":"https://openalex.org/keywords/reduction","display_name":"Reduction (mathematics)","score":0.350600004196167}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7651000022888184},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7297000288963318},{"id":"https://openalex.org/C148730421","wikidata":"https://www.wikidata.org/wiki/Q141090","display_name":"Encryption","level":2,"score":0.7131999731063843},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.6341999769210815},{"id":"https://openalex.org/C158338273","wikidata":"https://www.wikidata.org/wiki/Q2154943","display_name":"Homomorphic encryption","level":3,"score":0.5807999968528748},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5146999955177307},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.4221000075340271},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41499999165534973},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.357699990272522},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.35749998688697815},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.350600004196167},{"id":"https://openalex.org/C129916263","wikidata":"https://www.wikidata.org/wiki/Q1141183","display_name":"Backward chaining","level":4,"score":0.34630000591278076},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.2973000109195709},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2969000041484833},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.288100004196167},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.27900001406669617},{"id":"https://openalex.org/C9417928","wikidata":"https://www.wikidata.org/wiki/Q1070689","display_name":"Image processing","level":3,"score":0.2773999869823456},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.26820001006126404},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.265500009059906},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.2517000138759613}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.16834","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.16834","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.16834","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.16834","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":"article"},"sustainable_development_goals":[{"score":0.638110876083374,"id":"https://metadata.un.org/sdg/8","display_name":"Decent work and economic growth"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Privacy-preserving":[0],"machine":[1],"learning":[2],"(PPML)":[3],"has":[4],"become":[5],"increasingly":[6],"important":[7],"in":[8,157],"applications":[9],"where":[10],"sensitive":[11],"data":[12],"must":[13],"remain":[14],"confidential.":[15],"Homomorphic":[16],"Encryption":[17],"(HE)":[18],"enables":[19],"computation":[20],"directly":[21],"on":[22,39,109,178],"encrypted":[23,44,49,110,136,183],"data,":[24],"allowing":[25],"neural":[26,77],"network":[27],"inference":[28,40,58,123,174],"without":[29],"revealing":[30],"raw":[31],"inputs.":[32],"While":[33],"prior":[34],"works":[35],"have":[36],"largely":[37],"focused":[38],"over":[41],"a":[42,82,132,139,149,154,179],"single":[43],"image,":[45],"batch":[46,92,133,180],"processing":[47,131],"of":[48,125,134,143,176,181,187],"inputs":[50],"lags":[51],"behind,":[52],"despite":[53],"being":[54],"critical":[55],"for":[56,74,90],"high-throughput":[57],"scenarios":[59],"and":[60,99,106,112,153],"training-oriented":[61],"workloads.":[62],"In":[63],"this":[64,68],"work,":[65],"we":[66,170],"address":[67],"gap":[69],"by":[70],"developing":[71],"optimized":[72],"algorithms":[73,98],"batched":[75],"HE-friendly":[76,104],"networks.":[78],"We":[79,95],"also":[80],"introduced":[81],"pipeline":[83],"architecture":[84],"designed":[85],"to":[86,161],"maximize":[87],"resource":[88],"efficiency":[89],"different":[91],"size":[93],"execution.":[94],"implemented":[96],"these":[97],"evaluated":[100],"our":[101,118],"work":[102],"using":[103,185],"ResNet-20":[105],"ResNet-34":[107,168],"models":[108],"CIFAR-10":[111],"CIFAR-100":[113],"datasets,":[114],"respectively.":[115],"For":[116,165],"ResNet-20,":[117],"approach":[119],"achieves":[120],"an":[121,172],"amortized":[122,173],"time":[124,175],"8.86":[126],"seconds":[127],"per":[128],"image":[129],"when":[130],"512":[135],"images,":[137],"with":[138],"peak":[140],"memory":[141,158],"usage":[142,159],"98.96":[144],"GB.":[145],"These":[146],"results":[147],"represent":[148],"1.78x":[150],"runtime":[151],"improvement":[152],"3.74x":[155],"reduction":[156],"compared":[160],"the":[162,166],"state-of-the-art":[163],"design.":[164],"deeper":[167],"model,":[169],"achieve":[171],"28.14":[177],"256":[182],"images":[184],"246.78GB":[186],"RAM":[188]},"counts_by_year":[],"updated_date":"2026-04-22T06:07:44.442478","created_date":"2026-04-22T00:00:00"}