{"id":"https://openalex.org/W4224918069","doi":"https://doi.org/10.1109/icassp43922.2022.9747552","title":"Integer-Only Zero-Shot Quantization for Efficient Speech Recognition","display_name":"Integer-Only Zero-Shot Quantization for Efficient Speech Recognition","publication_year":2022,"publication_date":"2022-04-27","ids":{"openalex":"https://openalex.org/W4224918069","doi":"https://doi.org/10.1109/icassp43922.2022.9747552"},"language":"en","primary_location":{"id":"doi:10.1109/icassp43922.2022.9747552","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9747552","pdf_url":null,"source":{"id":"https://openalex.org/S4363607702","display_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5002255210","display_name":"Sehoon Kim","orcid":"https://orcid.org/0000-0002-9339-5480"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Sehoon Kim","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103894843","display_name":"Amir Gholami","orcid":null},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Amir Gholami","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102915910","display_name":"Zhewei Yao","orcid":"https://orcid.org/0000-0001-7678-4321"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhewei Yao","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020772870","display_name":"Nick Lee","orcid":"https://orcid.org/0000-0002-6209-0262"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nicholas Lee","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064872913","display_name":"Patrick Wang","orcid":"https://orcid.org/0000-0003-3117-8189"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Patrick Wang","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066429945","display_name":"Aniruddha Nrusimha","orcid":null},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Aniruddha Nrusimha","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004867366","display_name":"Bohan Zhai","orcid":null},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bohan Zhai","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089291480","display_name":"Tianren Gao","orcid":null},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Tianren Gao","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033006662","display_name":"Michael W. Mahoney","orcid":"https://orcid.org/0000-0001-7920-4652"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Michael W. Mahoney","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5047285420","display_name":"Kurt Keutzer","orcid":"https://orcid.org/0000-0003-3868-8501"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kurt Keutzer","raw_affiliation_strings":["University of California,Berkeley","University of California, Berkeley"],"affiliations":[{"raw_affiliation_string":"University of California,Berkeley","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California, Berkeley","institution_ids":["https://openalex.org/I95457486"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":10,"corresponding_author_ids":["https://openalex.org/A5002255210"],"corresponding_institution_ids":["https://openalex.org/I95457486"],"apc_list":null,"apc_paid":null,"fwci":2.0788,"has_fulltext":false,"cited_by_count":20,"citation_normalized_percentile":{"value":0.8889248,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":97,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"4288","last_page":"4292"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998000264167786,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998000264167786,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9921000003814697,"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/T11309","display_name":"Music and Audio Processing","score":0.9919999837875366,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/quantization","display_name":"Quantization (signal processing)","score":0.8333590030670166},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7410784363746643},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.7254742980003357},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5695540904998779},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5195265412330627},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.5140851140022278},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.4658653736114502},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.4257453680038452},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.402123361825943},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3575406074523926},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.1771862804889679}],"concepts":[{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.8333590030670166},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7410784363746643},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.7254742980003357},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5695540904998779},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5195265412330627},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.5140851140022278},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.4658653736114502},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.4257453680038452},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.402123361825943},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3575406074523926},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.1771862804889679},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp43922.2022.9747552","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9747552","pdf_url":null,"source":{"id":"https://openalex.org/S4363607702","display_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.5899999737739563}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":35,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W1522301498","https://openalex.org/W2889339831","https://openalex.org/W2936774411","https://openalex.org/W2963122961","https://openalex.org/W2973215447","https://openalex.org/W2982802130","https://openalex.org/W2998183051","https://openalex.org/W3007625080","https://openalex.org/W3015537910","https://openalex.org/W3016010032","https://openalex.org/W3024366773","https://openalex.org/W3034940165","https://openalex.org/W3034997257","https://openalex.org/W3035460915","https://openalex.org/W3095173472","https://openalex.org/W3097297321","https://openalex.org/W3097522836","https://openalex.org/W3097777922","https://openalex.org/W3097891912","https://openalex.org/W3137147200","https://openalex.org/W3170233084","https://openalex.org/W3171689473","https://openalex.org/W3175169239","https://openalex.org/W3177265267","https://openalex.org/W4287756219","https://openalex.org/W6629717138","https://openalex.org/W6631190155","https://openalex.org/W6761111670","https://openalex.org/W6774500937","https://openalex.org/W6779648490","https://openalex.org/W6784825895","https://openalex.org/W6786271619","https://openalex.org/W6788001715","https://openalex.org/W6791940793"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W98480971","https://openalex.org/W17155033","https://openalex.org/W2150291671","https://openalex.org/W2027972911","https://openalex.org/W2146343568","https://openalex.org/W2097707447","https://openalex.org/W3009327594","https://openalex.org/W2803935332"],"abstract_inverted_index":{"End-to-end":[0],"neural":[1],"network":[2],"models":[3,16,47,123],"achieve":[4,155],"improved":[5],"performance":[6],"on":[7,19,44,160],"various":[8],"automatic":[9],"speech":[10],"recognition":[11],"(ASR)":[12],"tasks.":[13],"However,":[14],"these":[15,92],"perform":[17],"poorly":[18],"edge":[20],"hardware":[21],"due":[22,84],"to":[23,35,85,121,130,143,157],"large":[24],"memory":[25],"and":[26,60,117,134,136,164],"computation":[27],"requirements.":[28],"While":[29],"quantizing":[30,45],"model":[31],"weights":[32],"and/or":[33,74],"activations":[34],"low-precision":[36],"can":[37],"be":[38,82],"a":[39,161,169],"promising":[40],"solution,":[41],"previous":[42,53],"research":[43],"ASR":[46,102],"is":[48],"limited.":[49],"In":[50,104],"particular,":[51,105],"the":[52,114,144],"approaches":[54],"use":[55,119],"floating-point":[56],"arithmetic":[57],"during":[58,77,124],"inference":[59],"thus":[61],"they":[62,71],"cannot":[63],"fully":[64],"exploit":[65],"efficient":[66],"integer":[67],"processing":[68],"units.":[69],"Moreover,":[70,153],"require":[72],"training":[73],"validation":[75],"data":[76,109],"quantization,":[78],"which":[79],"may":[80],"not":[81],"available":[83],"security":[86],"or":[87],"privacy":[88],"concerns.":[89],"To":[90],"address":[91],"limitations,":[93],"we":[94,106,118,154],"propose":[95],"an":[96],"integer-only,":[97],"zeroshot":[98],"quantization":[99],"scheme":[100],"for":[101],"models.":[103],"generate":[107],"synthetic":[108],"whose":[110],"runtime":[111],"statistics":[112],"resemble":[113],"real":[115],"data,":[116],"it":[120],"calibrate":[122],"quantization.":[125,177],"We":[126],"apply":[127],"our":[128],"method":[129],"quantize":[131],"QuartzNet,":[132],"Jasper,":[133],"Conformer":[135],"show":[137],"negligible":[138],"WER":[139,171],"degradation":[140,172],"as":[141],"compared":[142],"full-precision":[145],"baseline":[146],"models,":[147],"even":[148],"without":[149],"using":[150],"any":[151],"data.":[152],"up":[156],"2.35\u00d7":[158],"speedup":[159],"T4":[162],"GPU":[163],"4\u00d7":[165],"compression":[166],"rate,":[167],"with":[168,175],"modest":[170],"of":[173],"<1%":[174],"INT8":[176]},"counts_by_year":[{"year":2025,"cited_by_count":7},{"year":2024,"cited_by_count":5},{"year":2023,"cited_by_count":4},{"year":2022,"cited_by_count":4}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}