{"id":"https://openalex.org/W4408347314","doi":"https://doi.org/10.1109/icassp49660.2025.10887819","title":"Task-Aware Unified Source Separation","display_name":"Task-Aware Unified Source Separation","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408347314","doi":"https://doi.org/10.1109/icassp49660.2025.10887819"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10887819","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10887819","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 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/A5079146015","display_name":"Kohei Saijo","orcid":null},"institutions":[{"id":"https://openalex.org/I4210159266","display_name":"Mitsubishi Electric (United States)","ror":"https://ror.org/053jnhe44","country_code":"US","type":"company","lineage":["https://openalex.org/I1306287861","https://openalex.org/I4210133125","https://openalex.org/I4210159266"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kohei Saijo","raw_affiliation_strings":["Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA","institution_ids":["https://openalex.org/I4210159266"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055580486","display_name":"Janek Ebbers","orcid":null},"institutions":[{"id":"https://openalex.org/I4210159266","display_name":"Mitsubishi Electric (United States)","ror":"https://ror.org/053jnhe44","country_code":"US","type":"company","lineage":["https://openalex.org/I1306287861","https://openalex.org/I4210133125","https://openalex.org/I4210159266"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Janek Ebbers","raw_affiliation_strings":["Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA","institution_ids":["https://openalex.org/I4210159266"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102812631","display_name":"Fran\u00e7ois G. Germain","orcid":"https://orcid.org/0000-0002-8973-5315"},"institutions":[{"id":"https://openalex.org/I4210159266","display_name":"Mitsubishi Electric (United States)","ror":"https://ror.org/053jnhe44","country_code":"US","type":"company","lineage":["https://openalex.org/I1306287861","https://openalex.org/I4210133125","https://openalex.org/I4210159266"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Fran\u00e7ois G. Germain","raw_affiliation_strings":["Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA","institution_ids":["https://openalex.org/I4210159266"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086940921","display_name":"Gordon Wichern","orcid":"https://orcid.org/0000-0002-8597-6795"},"institutions":[{"id":"https://openalex.org/I4210159266","display_name":"Mitsubishi Electric (United States)","ror":"https://ror.org/053jnhe44","country_code":"US","type":"company","lineage":["https://openalex.org/I1306287861","https://openalex.org/I4210133125","https://openalex.org/I4210159266"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gordon Wichern","raw_affiliation_strings":["Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA","institution_ids":["https://openalex.org/I4210159266"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5064097430","display_name":"Jonathan Le Roux","orcid":"https://orcid.org/0000-0002-0158-2837"},"institutions":[{"id":"https://openalex.org/I4210159266","display_name":"Mitsubishi Electric (United States)","ror":"https://ror.org/053jnhe44","country_code":"US","type":"company","lineage":["https://openalex.org/I1306287861","https://openalex.org/I4210133125","https://openalex.org/I4210159266"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jonathan Le Roux","raw_affiliation_strings":["Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mitsubishi Electric Research Laboratories (MERL),Cambridge,MA,USA","institution_ids":["https://openalex.org/I4210159266"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":5.3201,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.95356944,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"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":0.9901999831199646,"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":0.9901999831199646,"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":0.9750000238418579,"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/T10711","display_name":"Target Tracking and Data Fusion in Sensor Networks","score":0.9017000198364258,"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":0.7753884792327881},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.6377907395362854},{"id":"https://openalex.org/keywords/separation","display_name":"Separation (statistics)","score":0.4773711860179901},{"id":"https://openalex.org/keywords/source-separation","display_name":"Source separation","score":0.4197095036506653},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.24510902166366577},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.1140529215335846},{"id":"https://openalex.org/keywords/systems-engineering","display_name":"Systems engineering","score":0.08165562152862549},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.07754039764404297}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7753884792327881},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6377907395362854},{"id":"https://openalex.org/C2776061190","wikidata":"https://www.wikidata.org/wiki/Q7451805","display_name":"Separation (statistics)","level":2,"score":0.4773711860179901},{"id":"https://openalex.org/C2776864781","wikidata":"https://www.wikidata.org/wiki/Q52617913","display_name":"Source separation","level":2,"score":0.4197095036506653},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.24510902166366577},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.1140529215335846},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.08165562152862549},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.07754039764404297}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10887819","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10887819","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":43,"referenced_works":["https://openalex.org/W2094721231","https://openalex.org/W2221409856","https://openalex.org/W2460742184","https://openalex.org/W2951130829","https://openalex.org/W2952218014","https://openalex.org/W2962753171","https://openalex.org/W2962866211","https://openalex.org/W2972411915","https://openalex.org/W2972541922","https://openalex.org/W2973062255","https://openalex.org/W2990666817","https://openalex.org/W2998657200","https://openalex.org/W3015199127","https://openalex.org/W3095263845","https://openalex.org/W3097906045","https://openalex.org/W3160050861","https://openalex.org/W3161934504","https://openalex.org/W3163652268","https://openalex.org/W3185109982","https://openalex.org/W3205879560","https://openalex.org/W4205689591","https://openalex.org/W4210493250","https://openalex.org/W4224871700","https://openalex.org/W4225300878","https://openalex.org/W4226151502","https://openalex.org/W4232282348","https://openalex.org/W4296069305","https://openalex.org/W4297841626","https://openalex.org/W4310873168","https://openalex.org/W4372346433","https://openalex.org/W4375928773","https://openalex.org/W4385756463","https://openalex.org/W4391021717","https://openalex.org/W4391021752","https://openalex.org/W4392903379","https://openalex.org/W4392903555","https://openalex.org/W4394896659","https://openalex.org/W4402111568","https://openalex.org/W4402112079","https://openalex.org/W4403126475","https://openalex.org/W6757817989","https://openalex.org/W6773772901","https://openalex.org/W6786621524"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2071676784","https://openalex.org/W2376932109","https://openalex.org/W2077498359"],"abstract_inverted_index":{"Several":[0],"attempts":[1],"have":[2,85],"been":[3],"made":[4],"to":[5,65,86,120,124,137,177],"handle":[6,138],"multiple":[7],"source":[8,21,27,107,123],"separation":[9,22,28,68,100,108,142,160],"tasks":[10,69,143,161],"such":[11,63],"as":[12],"speech":[13,15],"enhancement,":[14],"separation,":[16,19],"sound":[17,45],"event":[18],"music":[20],"(MSS),":[23],"or":[24,44],"cinematic":[25],"audio":[26,168],"(CASS)":[29],"with":[30],"a":[31,52,104,114],"single":[32],"model.":[33,110],"These":[34],"models":[35,64],"are":[36,74,79],"trained":[37],"on":[38,131,190],"large-scale":[39],"data":[40],"including":[41,144,170],"speech,":[42],"instruments,":[43],"events":[46],"and":[47,95,126,174],"can":[48],"often":[49],"successfully":[50,155],"separate":[51],"wide":[53],"range":[54],"of":[55,72,117],"sources.":[56],"However,":[57],"it":[58,136],"is":[59],"still":[60],"challenging":[61],"for":[62],"cover":[66],"all":[67,97,139],"because":[70],"some":[71,167],"them":[73],"contradictory":[75,145],"(e.g.,":[76],"musical":[77],"instruments":[78],"separated":[80],"in":[81,89],"MSS":[82],"while":[83],"they":[84],"be":[87],"grouped":[88],"CASS).":[90],"To":[91],"overcome":[92],"this":[93],"issue":[94],"support":[96],"the":[98,132,140,151,157,181,191],"major":[99,141,159],"tasks,":[101],"we":[102],"propose":[103],"task-aware":[105],"unified":[106],"(TUSS)":[109],"The":[111],"model":[112,154,183],"uses":[113],"variable":[115],"number":[116],"learnable":[118],"prompts":[119],"specify":[121],"which":[122],"separate,":[125],"changes":[127,184],"its":[128,185],"behavior":[129,186],"depending":[130,189],"given":[133],"prompts,":[134],"enabling":[135],"ones.":[146],"Experimental":[147],"results":[148],"demonstrate":[149,178],"that":[150],"proposed":[152],"TUSS":[153,182],"handles":[156],"five":[158],"mentioned":[162],"earlier.":[163],"We":[164],"also":[165],"provide":[166],"examples,":[169],"both":[171],"synthetic":[172],"mixtures":[173],"real":[175],"recordings,":[176],"how":[179],"flexibly":[180],"at":[187],"inference":[188],"prompts.":[192]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":3}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}