{"id":"https://openalex.org/W7148489985","doi":"https://doi.org/10.1109/asru65441.2025.11433846","title":"Spiralformer: Low Latency Encoder for Streaming Speech Recognition with Circular Layer Skipping and Early Exiting","display_name":"Spiralformer: Low Latency Encoder for Streaming Speech Recognition with Circular Layer Skipping and Early Exiting","publication_year":2025,"publication_date":"2025-12-06","ids":{"openalex":"https://openalex.org/W7148489985","doi":"https://doi.org/10.1109/asru65441.2025.11433846"},"language":null,"primary_location":{"id":"doi:10.1109/asru65441.2025.11433846","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asru65441.2025.11433846","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)","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/A5011937407","display_name":"Emiru Tsunoo","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Emiru Tsunoo","raw_affiliation_strings":["Sony Group Corporation,Japan"],"affiliations":[{"raw_affiliation_string":"Sony Group Corporation,Japan","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034216052","display_name":"Hayato Futami","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hayato Futami","raw_affiliation_strings":["Sony Group Corporation,Japan"],"affiliations":[{"raw_affiliation_string":"Sony Group Corporation,Japan","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109644502","display_name":"Yosuke Kashiwagi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yosuke Kashiwagi","raw_affiliation_strings":["Sony Group Corporation,Japan"],"affiliations":[{"raw_affiliation_string":"Sony Group Corporation,Japan","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047892839","display_name":"Siddhant Arora","orcid":"https://orcid.org/0000-0003-0375-496X"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Siddhant Arora","raw_affiliation_strings":["Carnegie Mellon University,USA"],"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University,USA","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5132789061","display_name":"Shinji Watanabe","orcid":null},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shinji Watanabe","raw_affiliation_strings":["Carnegie Mellon University,USA"],"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University,USA","institution_ids":["https://openalex.org/I74973139"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5011937407"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.75303311,"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":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.5202999711036682,"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":0.5202999711036682,"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":0.34860000014305115,"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/T10901","display_name":"Advanced Data Compression Techniques","score":0.01269999984651804,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/latency","display_name":"Latency (audio)","score":0.4733000099658966},{"id":"https://openalex.org/keywords/layer","display_name":"Layer (electronics)","score":0.45809999108314514},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.41850000619888306},{"id":"https://openalex.org/keywords/voice-activity-detection","display_name":"Voice activity detection","score":0.34279999136924744},{"id":"https://openalex.org/keywords/speech-processing","display_name":"Speech processing","score":0.31220000982284546}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6836000084877014},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6008999943733215},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.4733000099658966},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.45809999108314514},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.41850000619888306},{"id":"https://openalex.org/C204201278","wikidata":"https://www.wikidata.org/wiki/Q1332614","display_name":"Voice activity detection","level":3,"score":0.34279999136924744},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32429999113082886},{"id":"https://openalex.org/C61328038","wikidata":"https://www.wikidata.org/wiki/Q3358061","display_name":"Speech processing","level":2,"score":0.31220000982284546},{"id":"https://openalex.org/C46637626","wikidata":"https://www.wikidata.org/wiki/Q6693015","display_name":"Low latency (capital markets)","level":2,"score":0.2987000048160553},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2924000024795532},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.2906999886035919},{"id":"https://openalex.org/C509933004","wikidata":"https://www.wikidata.org/wiki/Q194163","display_name":"Broadband","level":2,"score":0.25099998712539673}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/asru65441.2025.11433846","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asru65441.2025.11433846","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)","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":30,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W2127141656","https://openalex.org/W2143612262","https://openalex.org/W2913718171","https://openalex.org/W2962677625","https://openalex.org/W2962784628","https://openalex.org/W2963211739","https://openalex.org/W2963414781","https://openalex.org/W3007433671","https://openalex.org/W3015457435","https://openalex.org/W3015927303","https://openalex.org/W3015974384","https://openalex.org/W3016010032","https://openalex.org/W3096524176","https://openalex.org/W3097777922","https://openalex.org/W3136363192","https://openalex.org/W3142589538","https://openalex.org/W3162665866","https://openalex.org/W3163907627","https://openalex.org/W3196784225","https://openalex.org/W3197148831","https://openalex.org/W4225277656","https://openalex.org/W4319862255","https://openalex.org/W4372347449","https://openalex.org/W4375869369","https://openalex.org/W4385245566","https://openalex.org/W4392904348","https://openalex.org/W4402111231","https://openalex.org/W4402111955","https://openalex.org/W4402112452"],"related_works":[],"abstract_inverted_index":{"For":[0],"streaming":[1],"speech":[2],"recognition,":[3],"a":[4,44,47,70,90],"Transformerbased":[5],"encoder":[6],"has":[7,25],"been":[8,26],"widely":[9],"used":[10],"with":[11,46,63,135,138],"block":[12,34,76,100,111],"processing.":[13,35,112],"Although":[14],"many":[15],"studies":[16],"addressed":[17],"improving":[18,29],"emission":[19,126],"latency":[20,31,40],"of":[21,32],"transducers,":[22],"little":[23],"work":[24],"explored":[27],"for":[28,75,105],"encoding":[30],"the":[33,64,95,107,110,123,136],"We":[36,85],"seek":[37],"to":[38],"reduce":[39],"by":[41,78],"frequently":[42],"emitting":[43],"chunk":[45,66],"small":[48,65],"shift":[49,94],"rather":[50],"than":[51],"scarce":[52],"large-chunk":[53],"emissions,":[54],"resulting":[55],"in":[56,89,98,122,128,132],"higher":[57],"computational":[58,140],"costs.":[59],"To":[60],"efficiently":[61],"compute":[62],"shift,":[67],"we":[68,114],"propose":[69],"new":[71],"encoder,":[72],"Spiralformer,":[73],"tailored":[74],"processing":[77],"combining":[79],"layer":[80,87,97],"dropping":[81],"and":[82,93,130,142],"early":[83],"exiting.":[84],"skip":[86],"computation":[88,104],"cyclic":[91],"manner":[92],"computed":[96],"each":[99],"spirally,":[101],"which":[102],"completes":[103],"all":[106],"layers":[108],"over":[109],"Experimentally,":[113],"observed":[115],"that":[116],"our":[117],"method":[118],"achieved":[119],"21.6%":[120],"reduction":[121],"averaged":[124],"token":[125],"delay":[127],"Librispeech,":[129],"7.0%":[131],"CSJ,":[133],"compared":[134],"baseline":[137],"similar":[139],"cost":[141],"word":[143],"error":[144],"rates.":[145]},"counts_by_year":[],"updated_date":"2026-04-03T16:44:17.987007","created_date":"2026-04-03T00:00:00"}