{"id":"https://openalex.org/W4408354432","doi":"https://doi.org/10.1109/icassp49660.2025.10889092","title":"What Are They Doing? Joint Audio-Speech Co-Reasoning","display_name":"What Are They Doing? Joint Audio-Speech Co-Reasoning","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408354432","doi":"https://doi.org/10.1109/icassp49660.2025.10889092"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10889092","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10889092","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/A5104907979","display_name":"Yingzhi Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210096041","display_name":"Riyadh Elm University","ror":"https://ror.org/00rz3mr26","country_code":"SA","type":"education","lineage":["https://openalex.org/I4210096041"]}],"countries":["SA"],"is_corresponding":true,"raw_author_name":"Yingzhi Wang","raw_affiliation_strings":["Elm,Research Center,Riyadh,KSA"],"affiliations":[{"raw_affiliation_string":"Elm,Research Center,Riyadh,KSA","institution_ids":["https://openalex.org/I4210096041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5038687274","display_name":"Pooneh Mousavi","orcid":"https://orcid.org/0009-0007-5761-5988"},"institutions":[{"id":"https://openalex.org/I60158472","display_name":"Concordia University","ror":"https://ror.org/0420zvk78","country_code":"CA","type":"education","lineage":["https://openalex.org/I60158472"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Pooneh Mousavi","raw_affiliation_strings":["Concordia University Mila,Montreal,Canada"],"affiliations":[{"raw_affiliation_string":"Concordia University Mila,Montreal,Canada","institution_ids":["https://openalex.org/I60158472"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063047435","display_name":"Artem Ploujnikov","orcid":null},"institutions":[{"id":"https://openalex.org/I4210164802","display_name":"Mila - Quebec Artificial Intelligence Institute","ror":"https://ror.org/05c22rx21","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210164802"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Artem Ploujnikov","raw_affiliation_strings":["Université de Montréal Mila,Montreal,Canada"],"affiliations":[{"raw_affiliation_string":"Université de Montréal Mila,Montreal,Canada","institution_ids":["https://openalex.org/I4210164802"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5040811098","display_name":"Mirco Ravanelli","orcid":"https://orcid.org/0000-0002-3929-5526"},"institutions":[{"id":"https://openalex.org/I60158472","display_name":"Concordia University","ror":"https://ror.org/0420zvk78","country_code":"CA","type":"education","lineage":["https://openalex.org/I60158472"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Mirco Ravanelli","raw_affiliation_strings":["Concordia University Mila,Montreal,Canada"],"affiliations":[{"raw_affiliation_string":"Concordia University Mila,Montreal,Canada","institution_ids":["https://openalex.org/I60158472"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5104907979"],"corresponding_institution_ids":["https://openalex.org/I4210096041"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.02064046,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"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/T12031","display_name":"Speech and dialogue systems","score":0.9491000175476074,"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":0.9491000175476074,"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/joint","display_name":"Joint (building)","score":0.7336000204086304},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7166436910629272},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.567625105381012},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.09765040874481201}],"concepts":[{"id":"https://openalex.org/C18555067","wikidata":"https://www.wikidata.org/wiki/Q8375051","display_name":"Joint (building)","level":2,"score":0.7336000204086304},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7166436910629272},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.567625105381012},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.09765040874481201},{"id":"https://openalex.org/C170154142","wikidata":"https://www.wikidata.org/wiki/Q150737","display_name":"Architectural engineering","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10889092","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10889092","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":33,"referenced_works":["https://openalex.org/W1649407914","https://openalex.org/W1977067929","https://openalex.org/W2074788634","https://openalex.org/W2081074144","https://openalex.org/W2118682956","https://openalex.org/W2127566601","https://openalex.org/W2129244720","https://openalex.org/W2593116425","https://openalex.org/W3178592608","https://openalex.org/W3209984917","https://openalex.org/W4391021627","https://openalex.org/W4391021638","https://openalex.org/W4392902623","https://openalex.org/W4402684170","https://openalex.org/W4404515031","https://openalex.org/W4404781855","https://openalex.org/W4411119642","https://openalex.org/W6732646663","https://openalex.org/W6733440450","https://openalex.org/W6746278845","https://openalex.org/W6796581206","https://openalex.org/W6800751262","https://openalex.org/W6809646742","https://openalex.org/W6846029027","https://openalex.org/W6847363464","https://openalex.org/W6848208918","https://openalex.org/W6849177959","https://openalex.org/W6852874933","https://openalex.org/W6854866820","https://openalex.org/W6857054612","https://openalex.org/W6859099255","https://openalex.org/W6866575372","https://openalex.org/W6870775139"],"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/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"In":[0],"audio":[1,11,27,41,86],"and":[2,19,42,87],"speech":[3,13,21,43,88],"processing,":[4,89],"tasks":[5],"usually":[6],"focus":[7],"on":[8,120],"either":[9],"the":[10,25,113],"or":[12],"modality,":[14],"even":[15],"when":[16],"both":[17,94],"sounds":[18],"human":[20],"are":[22],"present":[23],"in":[24],"same":[26],"clip.":[28],"Recent":[29],"Auditory":[30],"Large":[31],"Language":[32],"Models":[33],"(ALLMs)":[34],"have":[35],"made":[36],"it":[37],"possible":[38],"to":[39,50,64],"process":[40],"simultaneously":[44],"within":[45],"a":[46,61,81,99],"single":[47],"model,":[48],"leading":[49],"further":[51],"considerations":[52],"of":[53],"joint":[54,71],"audio-speech":[55,72],"tasks.In":[56],"this":[57],"paper,":[58],"we":[59,75,108],"establish":[60],"novel":[62,82],"benchmark":[63],"investigate":[65],"how":[66],"well":[67],"ALLMs":[68],"can":[69],"perform":[70],"processing.":[73],"Specifically,":[74],"introduce":[76],"Joint":[77],"Audio-Speech":[78],"Co-Reasoning":[79],"(JASCO),":[80],"task":[83],"that":[84],"unifies":[85],"strictly":[90],"requiring":[91],"co-reasoning":[92],"across":[93],"modalities.":[95],"We":[96],"also":[97],"release":[98],"scene-reasoning":[100],"dataset":[101],"called":[102],"\"What":[103],"Are":[104],"They":[105],"Doing\".":[106],"Additionally,":[107],"provide":[109],"deeper":[110],"insights":[111],"into":[112],"models\u2019":[114],"behaviors":[115],"by":[116],"analyzing":[117],"their":[118],"dependence":[119],"each":[121],"modality.":[122]},"counts_by_year":[],"updated_date":"2025-12-21T01:58:51.020947","created_date":"2025-10-10T00:00:00"}