{"id":"https://openalex.org/W4319862637","doi":"https://doi.org/10.1109/slt54892.2023.10023097","title":"Untied Positional Encodings for Efficient Transformer-Based Speech Recognition","display_name":"Untied Positional Encodings for Efficient Transformer-Based Speech Recognition","publication_year":2023,"publication_date":"2023-01-09","ids":{"openalex":"https://openalex.org/W4319862637","doi":"https://doi.org/10.1109/slt54892.2023.10023097"},"language":"en","primary_location":{"id":"doi:10.1109/slt54892.2023.10023097","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt54892.2023.10023097","pdf_url":null,"source":{"id":"https://openalex.org/S4363605953","display_name":"2022 IEEE Spoken Language Technology Workshop (SLT)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 IEEE Spoken Language Technology Workshop (SLT)","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/A5004424826","display_name":"Lahiru Samarakoon","orcid":null},"institutions":[{"id":"https://openalex.org/I4210090344","display_name":"Fano Labs (China)","ror":"https://ror.org/008jsgk36","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210090344"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lahiru Samarakoon","raw_affiliation_strings":["Fano Labs,Hong Kong","Fano Labs, Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Fano Labs,Hong Kong","institution_ids":["https://openalex.org/I4210090344"]},{"raw_affiliation_string":"Fano Labs, Hong Kong","institution_ids":["https://openalex.org/I4210090344"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5042559717","display_name":"Ivan W. H. Fung","orcid":"https://orcid.org/0000-0002-4396-151X"},"institutions":[{"id":"https://openalex.org/I4210090344","display_name":"Fano Labs (China)","ror":"https://ror.org/008jsgk36","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210090344"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ivan Fung","raw_affiliation_strings":["Fano Labs,Hong Kong","Fano Labs, Hong Kong"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Fano Labs,Hong Kong","institution_ids":["https://openalex.org/I4210090344"]},{"raw_affiliation_string":"Fano Labs, Hong Kong","institution_ids":["https://openalex.org/I4210090344"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4210090344"],"apc_list":null,"apc_paid":null,"fwci":0.4356,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.54381125,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"108","last_page":"114"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":1.0,"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":1.0,"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/T10860","display_name":"Speech and Audio Processing","score":0.9991000294685364,"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/T11309","display_name":"Music and Audio Processing","score":0.9983999729156494,"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/computer-science","display_name":"Computer science","score":0.6645447611808777},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6613028049468994},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.6331171989440918},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.6169480085372925},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5155658721923828},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.5140253305435181},{"id":"https://openalex.org/keywords/convolution","display_name":"Convolution (computer science)","score":0.4911210238933563},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4727940261363983},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.45617079734802246},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.4451947808265686},{"id":"https://openalex.org/keywords/correlation","display_name":"Correlation","score":0.43683671951293945},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3122192621231079},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.16895374655723572},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.08234474062919617},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.0785720944404602}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6645447611808777},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6613028049468994},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.6331171989440918},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.6169480085372925},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5155658721923828},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.5140253305435181},{"id":"https://openalex.org/C45347329","wikidata":"https://www.wikidata.org/wiki/Q5166604","display_name":"Convolution (computer science)","level":3,"score":0.4911210238933563},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4727940261363983},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.45617079734802246},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.4451947808265686},{"id":"https://openalex.org/C117220453","wikidata":"https://www.wikidata.org/wiki/Q5172842","display_name":"Correlation","level":2,"score":0.43683671951293945},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3122192621231079},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.16895374655723572},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.08234474062919617},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.0785720944404602},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.0},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/slt54892.2023.10023097","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt54892.2023.10023097","pdf_url":null,"source":{"id":"https://openalex.org/S4363605953","display_name":"2022 IEEE Spoken Language Technology Workshop (SLT)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 IEEE Spoken Language Technology Workshop (SLT)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","score":0.5,"display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":43,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W2327501763","https://openalex.org/W2766219058","https://openalex.org/W2892009249","https://openalex.org/W2919624000","https://openalex.org/W2936774411","https://openalex.org/W2941814890","https://openalex.org/W2948981900","https://openalex.org/W2962780374","https://openalex.org/W2963925437","https://openalex.org/W2964110616","https://openalex.org/W2973072704","https://openalex.org/W3008191852","https://openalex.org/W3008898571","https://openalex.org/W3015468748","https://openalex.org/W3033529678","https://openalex.org/W3097777922","https://openalex.org/W3097882114","https://openalex.org/W3122515622","https://openalex.org/W3122890974","https://openalex.org/W3133264589","https://openalex.org/W3161022068","https://openalex.org/W3170261818","https://openalex.org/W3173365702","https://openalex.org/W3197976839","https://openalex.org/W4225293011","https://openalex.org/W4288089799","https://openalex.org/W4295838474","https://openalex.org/W4309793872","https://openalex.org/W4385245566","https://openalex.org/W4388979610","https://openalex.org/W6755406732","https://openalex.org/W6760601182","https://openalex.org/W6761563299","https://openalex.org/W6769627184","https://openalex.org/W6771626834","https://openalex.org/W6775034530","https://openalex.org/W6776048684","https://openalex.org/W6779068807","https://openalex.org/W6779163297","https://openalex.org/W6779902714","https://openalex.org/W6783944145","https://openalex.org/W6784237645"],"related_works":["https://openalex.org/W4390516098","https://openalex.org/W2181948922","https://openalex.org/W2384362569","https://openalex.org/W4205302943","https://openalex.org/W2119949815","https://openalex.org/W2561132942","https://openalex.org/W2142795561","https://openalex.org/W3155418658","https://openalex.org/W4243199227","https://openalex.org/W2379948177"],"abstract_inverted_index":{"Self-attention":[0],"has":[1],"become":[2],"a":[3,27,80,127],"vital":[4],"component":[5],"for":[6],"end-to-end":[7],"(E2E)":[8],"automatic":[9],"speech":[10],"recognition":[11],"(ASR).":[12],"Convolution-augmented":[13],"Transformer":[14,107],"(Conformer)":[15],"with":[16,65,109],"relative":[17,114],"positional":[18,28,50,67],"encoding":[19,29],"(RPE)":[20],"achieved":[21],"state-of-the-art":[22],"performance.":[23],"This":[24],"paper":[25],"proposes":[26],"(PE)":[30],"mechanism":[31],"called":[32,82],"Scaled":[33,110],"Untied":[34,111],"RPE":[35,112],"that":[36,87,106],"unties":[37],"the":[38,42,66,70,88],"feature-position":[39],"correlations":[40,48,51,64,68],"in":[41,118,124],"self-attention":[43],"computation,":[44],"and":[45,49,69,120],"computes":[46],"feature":[47,63],"separately":[52],"using":[53,79],"different":[54],"projection":[55],"matrices.":[56],"In":[57],"addition,":[58],"we":[59,85],"propose":[60],"to":[61,94,122],"scale":[62],"aggressiveness":[71],"of":[72,116],"this":[73],"multiplicative":[74],"interaction":[75],"can":[76,91],"be":[77,92],"configured":[78],"parameter":[81],"amplitude.":[83],"Moreover,":[84],"show":[86,105],"PE":[89],"matrix":[90],"sliced":[93],"reduce":[95],"model":[96],"parameters.":[97],"Our":[98],"results":[99],"on":[100],"National":[101],"Speech":[102],"Corpus":[103],"(NSC)":[104],"encoders":[108],"achieves":[113],"improvements":[115],"1.9%":[117],"accuracy":[119],"up":[121],"50.9%":[123],"latency":[125],"over":[126],"Conformer":[128],"baseline":[129],"respectively.":[130]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2023,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}