{"id":"https://openalex.org/W4408352488","doi":"https://doi.org/10.1109/icassp49660.2025.10888561","title":"Salmon: A Suite for Acoustic Language Model Evaluation","display_name":"Salmon: A Suite for Acoustic Language Model Evaluation","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408352488","doi":"https://doi.org/10.1109/icassp49660.2025.10888561"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10888561","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10888561","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":"conference-paper","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/A5069765023","display_name":"Gallil Maimon","orcid":"https://orcid.org/0000-0001-6400-3954"},"institutions":[{"id":"https://openalex.org/I197251160","display_name":"Hebrew University of Jerusalem","ror":"https://ror.org/03qxff017","country_code":"IL","type":"education","lineage":["https://openalex.org/I197251160"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Gallil Maimon","raw_affiliation_strings":["Hebrew University of Jerusalem,School of Computer Science & Eng.,Jerusalem,Israel"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hebrew University of Jerusalem,School of Computer Science & Eng.,Jerusalem,Israel","institution_ids":["https://openalex.org/I197251160"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114353549","display_name":"Amit Roth","orcid":null},"institutions":[{"id":"https://openalex.org/I197251160","display_name":"Hebrew University of Jerusalem","ror":"https://ror.org/03qxff017","country_code":"IL","type":"education","lineage":["https://openalex.org/I197251160"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Amit Roth","raw_affiliation_strings":["Hebrew University of Jerusalem,School of Computer Science & Eng.,Jerusalem,Israel"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hebrew University of Jerusalem,School of Computer Science & Eng.,Jerusalem,Israel","institution_ids":["https://openalex.org/I197251160"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5005191803","display_name":"Yossi Adi","orcid":"https://orcid.org/0000-0003-2237-3898"},"institutions":[{"id":"https://openalex.org/I197251160","display_name":"Hebrew University of Jerusalem","ror":"https://ror.org/03qxff017","country_code":"IL","type":"education","lineage":["https://openalex.org/I197251160"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Yossi Adi","raw_affiliation_strings":["Hebrew University of Jerusalem,School of Computer Science & Eng.,Jerusalem,Israel"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hebrew University of Jerusalem,School of Computer Science & Eng.,Jerusalem,Israel","institution_ids":["https://openalex.org/I197251160"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I197251160"],"apc_list":null,"apc_paid":null,"fwci":10.0381,"has_fulltext":false,"cited_by_count":6,"citation_normalized_percentile":{"value":0.98058554,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9864000082015991,"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":0.9864000082015991,"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":0.9427000284194946,"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":0.9031999707221985,"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/suite","display_name":"Suite","score":0.8337737321853638},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6822665929794312},{"id":"https://openalex.org/keywords/history","display_name":"History","score":0.06596407294273376},{"id":"https://openalex.org/keywords/archaeology","display_name":"Archaeology","score":0.06001102924346924}],"concepts":[{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.8337737321853638},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6822665929794312},{"id":"https://openalex.org/C95457728","wikidata":"https://www.wikidata.org/wiki/Q309","display_name":"History","level":0,"score":0.06596407294273376},{"id":"https://openalex.org/C166957645","wikidata":"https://www.wikidata.org/wiki/Q23498","display_name":"Archaeology","level":1,"score":0.06001102924346924}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10888561","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10888561","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":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.5699999928474426}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":45,"referenced_works":["https://openalex.org/W2747874407","https://openalex.org/W3140429000","https://openalex.org/W3197259906","https://openalex.org/W3198217962","https://openalex.org/W3209059054","https://openalex.org/W3215615641","https://openalex.org/W4205689591","https://openalex.org/W4307323391","https://openalex.org/W4307680525","https://openalex.org/W4375868953","https://openalex.org/W4381786045","https://openalex.org/W4385822336","https://openalex.org/W4385823130","https://openalex.org/W4387595589","https://openalex.org/W4389524018","https://openalex.org/W4389524060","https://openalex.org/W4392902623","https://openalex.org/W4392904805","https://openalex.org/W4401042284","https://openalex.org/W4401246677","https://openalex.org/W4402111497","https://openalex.org/W4402112110","https://openalex.org/W4402112381","https://openalex.org/W4402684170","https://openalex.org/W4404782909","https://openalex.org/W4404783093","https://openalex.org/W4409348167","https://openalex.org/W4411119642","https://openalex.org/W6772349387","https://openalex.org/W6790356757","https://openalex.org/W6810325043","https://openalex.org/W6847363464","https://openalex.org/W6852781825","https://openalex.org/W6853530687","https://openalex.org/W6853998256","https://openalex.org/W6854866820","https://openalex.org/W6857054612","https://openalex.org/W6859099255","https://openalex.org/W6861618871","https://openalex.org/W6864829915","https://openalex.org/W6868911249","https://openalex.org/W6868986815","https://openalex.org/W6869365648","https://openalex.org/W6873786637","https://openalex.org/W6917585676"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4231704780","https://openalex.org/W2083794993","https://openalex.org/W352609212","https://openalex.org/W4404692318","https://openalex.org/W4200340037","https://openalex.org/W1511772879","https://openalex.org/W4379115841"],"abstract_inverted_index":{"Speech":[0],"language":[1,128],"models":[2,14,129],"have":[3,15],"recently":[4],"demonstrated":[5],"great":[6],"potential":[7],"as":[8,32],"universal":[9],"speech":[10,127],"processing":[11],"systems.":[12],"Such":[13],"the":[16,20,80,83,91,115,134,144],"ability":[17],"to":[18,44,118],"model":[19,103],"rich":[21],"acoustic":[22,49],"information":[23],"existing":[24],"in":[25],"audio":[26],"signals,":[27],"beyond":[28],"spoken":[29,92],"content,":[30],"such":[31],"emotion,":[33],"background":[34,66],"noise,":[35,67],"etc.":[36],"Despite":[37],"this,":[38],"evaluation":[39,63],"benchmarks":[40,77],"which":[41],"evaluate":[42,79],"awareness":[43],"a":[45,61,96,102],"wide":[46],"range":[47],"of":[48,82,138],"aspects,":[50],"are":[51],"lacking.":[52],"To":[53],"help":[54],"bridge":[55],"this":[56],"gap,":[57],"we":[58],"introduce":[59],"SALMON,":[60,131],"novel":[62],"suite":[64],"encompassing":[65],"sentiment,":[68],"speaker":[69],"identity":[70],"and":[71,86,136,146],"room":[72],"impulse":[73],"response.":[74],"The":[75],"proposed":[76],"both":[78],"consistency":[81],"inspected":[84],"element":[85],"how":[87],"much":[88],"it":[89],"matches":[90],"text.":[93],"We":[94,124,142],"follow":[95],"modelling":[97],"based":[98],"approach,":[99],"measuring":[100],"whether":[101],"gives":[104],"correct":[105],"samples":[106],"higher":[107],"scores":[108],"than":[109],"incorrect":[110],"ones.":[111],"This":[112],"approach":[113],"makes":[114],"benchmark":[116],"fast":[117],"compute":[119],"even":[120],"for":[121],"large":[122],"models.":[123],"evaluated":[125,140],"several":[126],"on":[130],"thus":[132],"highlighting":[133],"strengths":[135],"weaknesses":[137],"each":[139],"method.":[141],"make":[143],"code":[145],"data":[147],"publicly":[148],"available":[149],"at":[150],"-":[151],"pages.cs.huji.ac.il/adiyoss-lab/salmon/.":[152]},"counts_by_year":[{"year":2026,"cited_by_count":4},{"year":2025,"cited_by_count":2}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}