{"id":"https://openalex.org/W4375869127","doi":"https://doi.org/10.1109/icassp49357.2023.10096615","title":"Gradient Remedy for Multi-Task Learning in End-to-End Noise-Robust Speech Recognition","display_name":"Gradient Remedy for Multi-Task Learning in End-to-End Noise-Robust Speech Recognition","publication_year":2023,"publication_date":"2023-05-05","ids":{"openalex":"https://openalex.org/W4375869127","doi":"https://doi.org/10.1109/icassp49357.2023.10096615"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49357.2023.10096615","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10096615","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 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/A5074544822","display_name":"Yu\u2010Chen Hu","orcid":"https://orcid.org/0000-0002-5055-3645"},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":true,"raw_author_name":"Yuchen Hu","raw_affiliation_strings":["Nanyang Technological University,Singapore","Nanyang Technological University, Singapore"],"affiliations":[{"raw_affiliation_string":"Nanyang Technological University,Singapore","institution_ids":["https://openalex.org/I172675005"]},{"raw_affiliation_string":"Nanyang Technological University, Singapore","institution_ids":["https://openalex.org/I172675005"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100418370","display_name":"Chen Chen","orcid":"https://orcid.org/0000-0001-7587-3314"},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Chen Chen","raw_affiliation_strings":["Nanyang Technological University,Singapore","Nanyang Technological University, Singapore"],"affiliations":[{"raw_affiliation_string":"Nanyang Technological University,Singapore","institution_ids":["https://openalex.org/I172675005"]},{"raw_affiliation_string":"Nanyang Technological University, Singapore","institution_ids":["https://openalex.org/I172675005"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5018741641","display_name":"Ruizhe Li","orcid":"https://orcid.org/0000-0003-2512-845X"},"institutions":[{"id":"https://openalex.org/I195460627","display_name":"University of Aberdeen","ror":"https://ror.org/016476m91","country_code":"GB","type":"education","lineage":["https://openalex.org/I195460627"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Ruizhe Li","raw_affiliation_strings":["University of Aberdeen,UK","University of Aberdeen, UK"],"affiliations":[{"raw_affiliation_string":"University of Aberdeen,UK","institution_ids":["https://openalex.org/I195460627"]},{"raw_affiliation_string":"University of Aberdeen, UK","institution_ids":["https://openalex.org/I195460627"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045966396","display_name":"Qiushi Zhu","orcid":"https://orcid.org/0000-0002-1196-7781"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qiushi Zhu","raw_affiliation_strings":["University of Science and Technology of China,China","University of Science and Technology of China, China"],"affiliations":[{"raw_affiliation_string":"University of Science and Technology of China,China","institution_ids":["https://openalex.org/I126520041"]},{"raw_affiliation_string":"University of Science and Technology of China, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5070872826","display_name":"Eng Siong Chng","orcid":"https://orcid.org/0000-0001-6257-7399"},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Eng Siong Chng","raw_affiliation_strings":["Nanyang Technological University,Singapore","Nanyang Technological University, Singapore"],"affiliations":[{"raw_affiliation_string":"Nanyang Technological University,Singapore","institution_ids":["https://openalex.org/I172675005"]},{"raw_affiliation_string":"Nanyang Technological University, Singapore","institution_ids":["https://openalex.org/I172675005"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5074544822"],"corresponding_institution_ids":["https://openalex.org/I172675005"],"apc_list":null,"apc_paid":null,"fwci":2.8528,"has_fulltext":false,"cited_by_count":14,"citation_normalized_percentile":{"value":0.91624413,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9998999834060669,"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"}},"topics":[{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9998999834060669,"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"}},{"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/T11309","display_name":"Music and Audio Processing","score":0.9980000257492065,"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/task","display_name":"Task (project management)","score":0.732418954372406},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7237685918807983},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.6925129890441895},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6568341255187988},{"id":"https://openalex.org/keywords/word-error-rate","display_name":"Word error rate","score":0.5235390663146973},{"id":"https://openalex.org/keywords/interference","display_name":"Interference (communication)","score":0.5070197582244873},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.42541757225990295},{"id":"https://openalex.org/keywords/speech-enhancement","display_name":"Speech enhancement","score":0.41898104548454285},{"id":"https://openalex.org/keywords/noise-reduction","display_name":"Noise reduction","score":0.28327324986457825},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.10246965289115906},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.08211639523506165},{"id":"https://openalex.org/keywords/channel","display_name":"Channel (broadcasting)","score":0.06795099377632141}],"concepts":[{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.732418954372406},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7237685918807983},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.6925129890441895},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6568341255187988},{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.5235390663146973},{"id":"https://openalex.org/C32022120","wikidata":"https://www.wikidata.org/wiki/Q797225","display_name":"Interference (communication)","level":3,"score":0.5070197582244873},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.42541757225990295},{"id":"https://openalex.org/C2776182073","wikidata":"https://www.wikidata.org/wiki/Q7575395","display_name":"Speech enhancement","level":3,"score":0.41898104548454285},{"id":"https://openalex.org/C163294075","wikidata":"https://www.wikidata.org/wiki/Q581861","display_name":"Noise reduction","level":2,"score":0.28327324986457825},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.10246965289115906},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.08211639523506165},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.06795099377632141},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.0},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49357.2023.10096615","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10096615","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 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/4","display_name":"Quality Education","score":0.47999998927116394}],"awards":[],"funders":[{"id":"https://openalex.org/F4320320671","display_name":"National Research Foundation","ror":"https://ror.org/05s0g1g46"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":33,"referenced_works":["https://openalex.org/W1522301498","https://openalex.org/W2024490156","https://openalex.org/W2064675550","https://openalex.org/W2250499536","https://openalex.org/W2895387432","https://openalex.org/W2963250244","https://openalex.org/W2998616931","https://openalex.org/W3032514799","https://openalex.org/W3097777922","https://openalex.org/W3106539628","https://openalex.org/W3120612350","https://openalex.org/W3142252347","https://openalex.org/W3183454191","https://openalex.org/W3206531472","https://openalex.org/W4226046733","https://openalex.org/W4297841258","https://openalex.org/W4312437435","https://openalex.org/W4372266968","https://openalex.org/W4372267073","https://openalex.org/W4372342491","https://openalex.org/W4375868977","https://openalex.org/W4375869103","https://openalex.org/W4382202533","https://openalex.org/W4385245566","https://openalex.org/W6631190155","https://openalex.org/W6656414902","https://openalex.org/W6691605950","https://openalex.org/W6739901393","https://openalex.org/W6771876938","https://openalex.org/W6846585435","https://openalex.org/W6846626936","https://openalex.org/W6850366547","https://openalex.org/W6996806968"],"related_works":["https://openalex.org/W4307321803","https://openalex.org/W2131674787","https://openalex.org/W2899416438","https://openalex.org/W2081983610","https://openalex.org/W2100534065","https://openalex.org/W4206328857","https://openalex.org/W2355125052","https://openalex.org/W3161525290","https://openalex.org/W4312822757","https://openalex.org/W2944686050"],"abstract_inverted_index":{"Speech":[0],"enhancement":[1],"(SE)":[2],"is":[3,23,119,196],"proved":[4],"effective":[5,83],"in":[6,95,126,136],"reducing":[7],"noise":[8],"from":[9,99,154],"noisy":[10],"speech":[11,16,34,97],"signals":[12],"for":[13],"downstream":[14],"automatic":[15],"recognition":[17],"(ASR),":[18],"where":[19],"multi-task":[20,66,185],"learning":[21,67,186],"strategy":[22],"employed":[24],"to":[25,71,89,123,128,148],"jointly":[26],"optimize":[27],"these":[28],"two":[29,146],"tasks.":[30],"However,":[31],"the":[32,47,55,65,110,130,143,150,164,169],"enhanced":[33],"learned":[35],"by":[36,157],"SE":[37,58,111,158],"objective":[38],"may":[39],"not":[40],"always":[41],"yield":[42],"good":[43],"ASR":[44,60,73,124,137,152],"results.":[45],"From":[46],"optimization":[48],"view,":[49],"there":[50],"sometimes":[51],"exists":[52],"interference":[53,91,171],"between":[54,92,132],"gradients":[56,94,147],"of":[57,101,145,180],"and":[59,68,104,134,172,182,190],"tasks,":[61],"which":[62],"could":[63],"hinder":[64],"finally":[69],"lead":[70],"sub-optimal":[72],"performance.":[74],"In":[75],"this":[76],"paper,":[77],"we":[78,107,140],"propose":[79],"a":[80,115],"simple":[81],"yet":[82],"approach":[84,166],"called":[85],"gradient":[86,113,170],"remedy":[87],"(GR)":[88],"solve":[90],"task":[93,153],"noise-robust":[96],"recognition,":[98],"perspectives":[100],"both":[102],"angle":[103,122],"magnitude.":[105],"Specifically,":[106],"first":[108],"project":[109],"task's":[112],"onto":[114],"dynamic":[116],"surface":[117],"that":[118,163],"at":[120,198],"acute":[121],"gradient,":[125],"order":[127],"remove":[129],"conflict":[131],"them":[133],"assist":[135],"optimization.":[138],"Furthermore,":[139],"adaptively":[141],"rescale":[142],"magnitude":[144],"prevent":[149],"dominant":[151],"being":[155],"misled":[156],"gradient.":[159],"Experimental":[160],"results":[161],"show":[162],"proposed":[165],"well":[167],"resolves":[168],"achieves":[173],"relative":[174],"word":[175],"error":[176],"rate":[177],"(WER)":[178],"reductions":[179],"9.3%":[181],"11.1%":[183],"over":[184],"baseline,":[187],"on":[188],"RATS":[189],"CHiME-4":[191],"datasets,":[192],"respectively.":[193],"Our":[194],"code":[195],"available":[197],"GitHub":[199],"<sup":[200],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[201],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">1</sup>":[202],".":[203]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":8},{"year":2023,"cited_by_count":4}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}