{"id":"https://openalex.org/W4392909582","doi":"https://doi.org/10.1109/icassp48485.2024.10446589","title":"Synthetic Conversations Improve Multi-Talker ASR","display_name":"Synthetic Conversations Improve Multi-Talker ASR","publication_year":2024,"publication_date":"2024-03-18","ids":{"openalex":"https://openalex.org/W4392909582","doi":"https://doi.org/10.1109/icassp48485.2024.10446589"},"language":"en","primary_location":{"id":"doi:10.1109/icassp48485.2024.10446589","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp48485.2024.10446589","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2024 - 2024 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/A5044152577","display_name":"Thai Binh Nguyen","orcid":"https://orcid.org/0009-0008-8100-6338"},"institutions":[{"id":"https://openalex.org/I102335020","display_name":"Karlsruhe Institute of Technology","ror":"https://ror.org/04t3en479","country_code":"DE","type":"education","lineage":["https://openalex.org/I102335020","https://openalex.org/I1305996414"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Thai-Binh Nguyen","raw_affiliation_strings":["Karlsruhe Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Karlsruhe Institute of Technology","institution_ids":["https://openalex.org/I102335020"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5023053982","display_name":"Alexander Waibel","orcid":null},"institutions":[{"id":"https://openalex.org/I102335020","display_name":"Karlsruhe Institute of Technology","ror":"https://ror.org/04t3en479","country_code":"DE","type":"education","lineage":["https://openalex.org/I102335020","https://openalex.org/I1305996414"]},{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["DE","US"],"is_corresponding":false,"raw_author_name":"Alexander Waibel","raw_affiliation_strings":["Karlsruhe Institute of Technology","Carnegie Mellon University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Karlsruhe Institute of Technology","institution_ids":["https://openalex.org/I102335020"]},{"raw_affiliation_string":"Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.5273,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.84476247,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"10461","last_page":"10465"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12031","display_name":"Speech and dialogue systems","score":0.9537000060081482,"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/T12031","display_name":"Speech and dialogue systems","score":0.9537000060081482,"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/computer-science","display_name":"Computer science","score":0.7039982080459595},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5066931843757629}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7039982080459595},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5066931843757629}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp48485.2024.10446589","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp48485.2024.10446589","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320309895","display_name":"Carl-Zeiss-Stiftung","ror":"https://ror.org/03ng4kg22"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":35,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W2125336414","https://openalex.org/W2159591770","https://openalex.org/W2251329024","https://openalex.org/W2460742184","https://openalex.org/W2799473636","https://openalex.org/W2973127116","https://openalex.org/W2979826702","https://openalex.org/W3016232124","https://openalex.org/W3027008958","https://openalex.org/W3034999214","https://openalex.org/W3094831814","https://openalex.org/W3094965760","https://openalex.org/W3095073040","https://openalex.org/W3101648800","https://openalex.org/W3119308075","https://openalex.org/W3143843080","https://openalex.org/W3157070662","https://openalex.org/W3181056718","https://openalex.org/W3198482860","https://openalex.org/W3198694222","https://openalex.org/W3198763610","https://openalex.org/W3203417382","https://openalex.org/W3212886388","https://openalex.org/W4311000453","https://openalex.org/W4372342485","https://openalex.org/W4372349800","https://openalex.org/W4378105483","https://openalex.org/W4385822321","https://openalex.org/W6631362777","https://openalex.org/W6771467084","https://openalex.org/W6777776875","https://openalex.org/W6847363464","https://openalex.org/W6852909395","https://openalex.org/W6898634591"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052","https://openalex.org/W2382290278","https://openalex.org/W4395014643"],"abstract_inverted_index":{"In":[0,40],"recent":[1],"times,":[2],"automatic":[3],"speech":[4,29],"recognition":[5],"(ASR)":[6],"has":[7],"seen":[8],"remarkable":[9],"progress,":[10],"particularly":[11],"in":[12,113],"recognizing":[13],"dominant":[14],"speakers.":[15],"Nevertheless,":[16],"the":[17,86,97,107,115,120],"challenge":[18],"of":[19,92,110,122],"multi-talker":[20,63,74,117],"scenarios":[21],"involving":[22],"distinguishing":[23],"between":[24],"speakers":[25],"and":[26,72,90,94,99],"transcribing":[27],"their":[28],"accurately":[30],"remains":[31],"unsolved":[32],"due":[33],"to":[34,58],"limited":[35],"data":[36],"constraining":[37],"model":[38,65],"effectiveness.":[39],"this":[41],"study,":[42],"We":[43],"propose":[44],"a":[45,79],"novel":[46],"methodology":[47],"called":[48],"Systematic":[49],"Synthetic":[50],"Conversations":[51],"(SSC),":[52],"which":[53],"leverages":[54],"conventional":[55],"ASR":[56,64],"datasets":[57],"help":[59],"an":[60],"end-to-end":[61],"(E2E)":[62],"establish":[66],"new":[67],"state-of-the-art":[68],"results":[69],"across":[70],"synthetic":[71],"authentic":[73],"datasets.":[75],"Notably,":[76],"we":[77],"achieved":[78],"3.47%":[80],"word":[81],"error":[82],"rate":[83],"(WER)":[84],"for":[85,96],"Libri2Mix":[87],"[1]":[88],"set,":[89],"WERs":[91],"13.96%":[93],"19.51%":[95],"AMI-IHM":[98],"AMI-SDM":[100],"[2]":[101],"sets,":[102],"respectively.":[103],"These":[104],"outcomes":[105],"underscore":[106],"hidden":[108],"potential":[109],"existing":[111],"resources":[112],"tackling":[114],"complicated":[116],"problems":[118],"within":[119],"domain":[121],"ASR.":[123]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":1}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}