{"id":"https://openalex.org/W4372259765","doi":"https://doi.org/10.1109/icassp49357.2023.10095168","title":"Effectiveness of Text, Acoustic, and Lattice-Based Representations in Spoken Language Understanding Tasks","display_name":"Effectiveness of Text, Acoustic, and Lattice-Based Representations in Spoken Language Understanding Tasks","publication_year":2023,"publication_date":"2023-05-05","ids":{"openalex":"https://openalex.org/W4372259765","doi":"https://doi.org/10.1109/icassp49357.2023.10095168"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49357.2023.10095168","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/icassp49357.2023.10095168","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/A5075765823","display_name":"Esa\u00fa Villatoro-Tello","orcid":"https://orcid.org/0000-0002-1322-0358"},"institutions":[{"id":"https://openalex.org/I7495430","display_name":"Idiap Research Institute","ror":"https://ror.org/05932h694","country_code":"CH","type":"facility","lineage":["https://openalex.org/I7495430"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Esa\u00fa Villatoro-Tello","raw_affiliation_strings":["Idiap Research Institute,Martigny,Switzerland","Idiap Research Institute, Martigny, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Idiap Research Institute,Martigny,Switzerland","institution_ids":["https://openalex.org/I7495430"]},{"raw_affiliation_string":"Idiap Research Institute, Martigny, Switzerland","institution_ids":["https://openalex.org/I7495430"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5084521938","display_name":"Srikanth Madikeri","orcid":"https://orcid.org/0000-0002-4361-784X"},"institutions":[{"id":"https://openalex.org/I7495430","display_name":"Idiap Research Institute","ror":"https://ror.org/05932h694","country_code":"CH","type":"facility","lineage":["https://openalex.org/I7495430"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Srikanth Madikeri","raw_affiliation_strings":["Idiap Research Institute,Martigny,Switzerland","Idiap Research Institute, Martigny, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Idiap Research Institute,Martigny,Switzerland","institution_ids":["https://openalex.org/I7495430"]},{"raw_affiliation_string":"Idiap Research Institute, Martigny, Switzerland","institution_ids":["https://openalex.org/I7495430"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078502662","display_name":"Juan Zuluaga-G\u00f3mez","orcid":"https://orcid.org/0000-0002-6947-2706"},"institutions":[{"id":"https://openalex.org/I5124864","display_name":"\u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne","ror":"https://ror.org/02s376052","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I5124864"]},{"id":"https://openalex.org/I7495430","display_name":"Idiap Research Institute","ror":"https://ror.org/05932h694","country_code":"CH","type":"facility","lineage":["https://openalex.org/I7495430"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Juan Zuluaga-Gomez","raw_affiliation_strings":["Idiap Research Institute,Martigny,Switzerland","Ecole Polytechnique Federale de Lausanne (EPFL), Switzerland","Idiap Research Institute, Martigny, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Idiap Research Institute,Martigny,Switzerland","institution_ids":["https://openalex.org/I7495430"]},{"raw_affiliation_string":"Ecole Polytechnique Federale de Lausanne (EPFL), Switzerland","institution_ids":["https://openalex.org/I5124864"]},{"raw_affiliation_string":"Idiap Research Institute, Martigny, Switzerland","institution_ids":["https://openalex.org/I7495430"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067433878","display_name":"Bidisha Sharma","orcid":"https://orcid.org/0000-0002-4195-3532"},"institutions":[{"id":"https://openalex.org/I4210099960","display_name":"Systems Control (United States)","ror":"https://ror.org/014tted12","country_code":"US","type":"company","lineage":["https://openalex.org/I4210099960"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bidisha Sharma","raw_affiliation_strings":["Uniphore Software Systems Inc.,Palo Alto,CA,USA","Uniphore Software Systems Inc., Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Uniphore Software Systems Inc.,Palo Alto,CA,USA","institution_ids":["https://openalex.org/I4210099960"]},{"raw_affiliation_string":"Uniphore Software Systems Inc., Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5085223859","display_name":"Seyyed Saeed Sarfjoo","orcid":null},"institutions":[{"id":"https://openalex.org/I7495430","display_name":"Idiap Research Institute","ror":"https://ror.org/05932h694","country_code":"CH","type":"facility","lineage":["https://openalex.org/I7495430"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Seyyed Saeed Sarfjoo","raw_affiliation_strings":["Idiap Research Institute,Martigny,Switzerland","Idiap Research Institute, Martigny, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Idiap Research Institute,Martigny,Switzerland","institution_ids":["https://openalex.org/I7495430"]},{"raw_affiliation_string":"Idiap Research Institute, Martigny, Switzerland","institution_ids":["https://openalex.org/I7495430"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021226188","display_name":"Iuliia Nigmatulina","orcid":"https://orcid.org/0000-0002-5111-1873"},"institutions":[{"id":"https://openalex.org/I7495430","display_name":"Idiap Research Institute","ror":"https://ror.org/05932h694","country_code":"CH","type":"facility","lineage":["https://openalex.org/I7495430"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Iuliia Nigmatulina","raw_affiliation_strings":["Idiap Research Institute,Martigny,Switzerland","Idiap Research Institute, Martigny, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Idiap Research Institute,Martigny,Switzerland","institution_ids":["https://openalex.org/I7495430"]},{"raw_affiliation_string":"Idiap Research Institute, Martigny, Switzerland","institution_ids":["https://openalex.org/I7495430"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076409146","display_name":"Petr Motl\u00ed\u010dek","orcid":"https://orcid.org/0000-0001-6467-1119"},"institutions":[{"id":"https://openalex.org/I60587646","display_name":"Brno University of Technology","ror":"https://ror.org/03613d656","country_code":"CZ","type":"education","lineage":["https://openalex.org/I60587646"]},{"id":"https://openalex.org/I7495430","display_name":"Idiap Research Institute","ror":"https://ror.org/05932h694","country_code":"CH","type":"facility","lineage":["https://openalex.org/I7495430"]}],"countries":["CH","CZ"],"is_corresponding":false,"raw_author_name":"Petr Motlicek","raw_affiliation_strings":["Idiap Research Institute,Martigny,Switzerland","Idiap Research Institute, Martigny, Switzerland","Brno University of Technology, Brno, Czech Republic"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Idiap Research Institute,Martigny,Switzerland","institution_ids":["https://openalex.org/I7495430"]},{"raw_affiliation_string":"Idiap Research Institute, Martigny, Switzerland","institution_ids":["https://openalex.org/I7495430"]},{"raw_affiliation_string":"Brno University of Technology, Brno, Czech Republic","institution_ids":["https://openalex.org/I60587646"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109479987","display_name":"Alexei V. Ivanov","orcid":null},"institutions":[{"id":"https://openalex.org/I4210099960","display_name":"Systems Control (United States)","ror":"https://ror.org/014tted12","country_code":"US","type":"company","lineage":["https://openalex.org/I4210099960"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Alexei V. Ivanov","raw_affiliation_strings":["Uniphore Software Systems Inc.,Palo Alto,CA,USA","Uniphore Software Systems Inc., Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Uniphore Software Systems Inc.,Palo Alto,CA,USA","institution_ids":["https://openalex.org/I4210099960"]},{"raw_affiliation_string":"Uniphore Software Systems Inc., Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5083410051","display_name":"Aravind Ganapathiraju","orcid":null},"institutions":[{"id":"https://openalex.org/I4210099960","display_name":"Systems Control (United States)","ror":"https://ror.org/014tted12","country_code":"US","type":"company","lineage":["https://openalex.org/I4210099960"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Aravind Ganapathiraju","raw_affiliation_strings":["Uniphore Software Systems Inc.,Palo Alto,CA,USA","Uniphore Software Systems Inc., Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Uniphore Software Systems Inc.,Palo Alto,CA,USA","institution_ids":["https://openalex.org/I4210099960"]},{"raw_affiliation_string":"Uniphore Software Systems Inc., Palo Alto, CA, USA","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":9,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.3263,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.6295477,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":94},"biblio":{"volume":"32","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":0.9998999834060669,"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":0.9998999834060669,"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/T12031","display_name":"Speech and dialogue systems","score":0.9997000098228455,"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":0.9990000128746033,"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":0.8617545366287231},{"id":"https://openalex.org/keywords/oracle","display_name":"Oracle","score":0.7087266445159912},{"id":"https://openalex.org/keywords/spoken-language","display_name":"Spoken language","score":0.6521936058998108},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.5635824799537659},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.4900604486465454},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4852485656738281},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.46080708503723145},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.4555460512638092}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8617545366287231},{"id":"https://openalex.org/C55166926","wikidata":"https://www.wikidata.org/wiki/Q2892946","display_name":"Oracle","level":2,"score":0.7087266445159912},{"id":"https://openalex.org/C2776230583","wikidata":"https://www.wikidata.org/wiki/Q1322198","display_name":"Spoken language","level":2,"score":0.6521936058998108},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5635824799537659},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.4900604486465454},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4852485656738281},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.46080708503723145},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.4555460512638092},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49357.2023.10095168","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/icassp49357.2023.10095168","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","score":0.75,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":40,"referenced_works":["https://openalex.org/W1524333225","https://openalex.org/W2077302143","https://openalex.org/W2189256702","https://openalex.org/W2251058040","https://openalex.org/W2618015847","https://openalex.org/W2889282842","https://openalex.org/W2894164357","https://openalex.org/W2896457183","https://openalex.org/W2963288440","https://openalex.org/W2964051877","https://openalex.org/W2972584841","https://openalex.org/W2979486337","https://openalex.org/W3008912312","https://openalex.org/W3015237657","https://openalex.org/W3036601975","https://openalex.org/W3095711142","https://openalex.org/W3096589040","https://openalex.org/W3096851323","https://openalex.org/W3097798370","https://openalex.org/W3100460087","https://openalex.org/W3163044982","https://openalex.org/W3167533889","https://openalex.org/W3198429080","https://openalex.org/W3198484663","https://openalex.org/W3209059054","https://openalex.org/W4205701539","https://openalex.org/W4284702538","https://openalex.org/W4287645199","https://openalex.org/W4295312788","https://openalex.org/W4297683418","https://openalex.org/W4385245566","https://openalex.org/W6631362777","https://openalex.org/W6739901393","https://openalex.org/W6751425476","https://openalex.org/W6755207826","https://openalex.org/W6766978945","https://openalex.org/W6779919476","https://openalex.org/W6780218876","https://openalex.org/W6784162623","https://openalex.org/W6791539774"],"related_works":["https://openalex.org/W2378211422","https://openalex.org/W2745001401","https://openalex.org/W4321353415","https://openalex.org/W2073713056","https://openalex.org/W2130974462","https://openalex.org/W972276598","https://openalex.org/W4246352526","https://openalex.org/W2028665553","https://openalex.org/W3110702597","https://openalex.org/W2086519370"],"abstract_inverted_index":{"In":[0],"this":[1],"paper,":[2],"we":[3],"perform":[4,31],"an":[5],"exhaustive":[6],"evaluation":[7],"of":[8,28,53,60,93,136,150],"different":[9,61,66],"representations":[10],"to":[11,30,105,109,129,146],"address":[12],"the":[13,32,57,75,78,102,110,130,139,148],"intent":[14,34],"classification":[15],"problem":[16],"in":[17,107],"a":[18,42,50,133,143],"Spoken":[19],"Language":[20],"Understanding":[21],"(SLU)":[22],"setup.":[23],"We":[24,73],"benchmark":[25],"three":[26],"types":[27],"systems":[29,64,76],"SLU":[33,63,103],"detection":[35],"task:":[36],"1)":[37],"text-based,":[38],"2)":[39],"lattice-based,":[40],"and":[41,123],"novel":[43],"3)":[44],"multimodal":[45],"approach.":[46],"Our":[47,86],"work":[48],"provides":[49],"comprehensive":[51],"analysis":[52],"what":[54],"could":[55],"be":[56],"achievable":[58],"performance":[59,127],"state-of-the-art":[62],"under":[65],"circumstances,":[67],"e.g.,":[68],"automatically-":[69],"vs.":[70],"manually-generated":[71],"transcripts.":[72,155],"evaluate":[74],"on":[77],"publicly":[79],"available":[80],"SLURP":[81],"spoken":[82],"language":[83],"resource":[84],"corpus.":[85],"results":[87],"indicate":[88],"that":[89],"using":[90],"richer":[91],"forms":[92],"Automatic":[94],"Speech":[95],"Recognition":[96],"(ASR)":[97],"outputs,":[98],"namely":[99],"word-consensus-networks,":[100],"allows":[101],"system":[104],"improve":[106],"comparison":[108],"1-best":[111,140],"setup":[112],"(5.5%":[113],"relative":[114,134],"improvement).":[115],"However,":[116],"crossmodal":[117],"approaches,":[118],"i.e.,":[119],"learning":[120],"from":[121],"acoustic":[122],"text":[124],"embeddings,":[125],"obtains":[126],"similar":[128],"oracle":[131],"setup,":[132],"improvement":[135],"17.8%":[137],"over":[138],"configuration,":[141],"being":[142],"recommended":[144],"alternative":[145],"overcome":[147],"limitations":[149],"working":[151],"with":[152],"automatically":[153],"generated":[154]},"counts_by_year":[{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":1}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}