{"id":"https://openalex.org/W4416799997","doi":"https://doi.org/10.1109/apsipaasc65261.2025.11249088","title":"Efficient Transformer-Based Piano Transcription with Sparse Attention Mechanisms","display_name":"Efficient Transformer-Based Piano Transcription with Sparse Attention Mechanisms","publication_year":2025,"publication_date":"2025-10-22","ids":{"openalex":"https://openalex.org/W4416799997","doi":"https://doi.org/10.1109/apsipaasc65261.2025.11249088"},"language":null,"primary_location":{"id":"doi:10.1109/apsipaasc65261.2025.11249088","is_oa":false,"landing_page_url":"https://doi.org/10.1109/apsipaasc65261.2025.11249088","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)","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/A5076273993","display_name":"Weixing Wei","orcid":null},"institutions":[{"id":"https://openalex.org/I22299242","display_name":"Kyoto University","ror":"https://ror.org/02kpeqv85","country_code":"JP","type":"education","lineage":["https://openalex.org/I22299242"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Weixing Wei","raw_affiliation_strings":["Kyoto University,Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Kyoto University,Japan","institution_ids":["https://openalex.org/I22299242"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5067956319","display_name":"Kazuyoshi Yoshii","orcid":"https://orcid.org/0000-0001-8387-8609"},"institutions":[{"id":"https://openalex.org/I22299242","display_name":"Kyoto University","ror":"https://ror.org/02kpeqv85","country_code":"JP","type":"education","lineage":["https://openalex.org/I22299242"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Kazuyoshi Yoshii","raw_affiliation_strings":["Kyoto University,Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Kyoto University,Japan","institution_ids":["https://openalex.org/I22299242"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I22299242"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.3882721,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"234","last_page":"239"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11309","display_name":"Music and Audio Processing","score":0.873199999332428,"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/T11309","display_name":"Music and Audio Processing","score":0.873199999332428,"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/T11349","display_name":"Music Technology and Sound Studies","score":0.0738999992609024,"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"}},{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.011599999852478504,"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/midi","display_name":"MIDI","score":0.7382000088691711},{"id":"https://openalex.org/keywords/piano","display_name":"Piano","score":0.676800012588501},{"id":"https://openalex.org/keywords/transcription","display_name":"Transcription (linguistics)","score":0.5799000263214111},{"id":"https://openalex.org/keywords/pooling","display_name":"Pooling","score":0.5656999945640564},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5090000033378601},{"id":"https://openalex.org/keywords/computational-complexity-theory","display_name":"Computational complexity theory","score":0.477400004863739},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.47040000557899475},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.4519999921321869},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.4235999882221222}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7835999727249146},{"id":"https://openalex.org/C8112396","wikidata":"https://www.wikidata.org/wiki/Q80535","display_name":"MIDI","level":2,"score":0.7382000088691711},{"id":"https://openalex.org/C124086623","wikidata":"https://www.wikidata.org/wiki/Q5994","display_name":"Piano","level":2,"score":0.676800012588501},{"id":"https://openalex.org/C179926584","wikidata":"https://www.wikidata.org/wiki/Q207714","display_name":"Transcription (linguistics)","level":2,"score":0.5799000263214111},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.5656999945640564},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5529999732971191},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5090000033378601},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.477400004863739},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.47040000557899475},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.462799996137619},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4519999921321869},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.4235999882221222},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.3075000047683716},{"id":"https://openalex.org/C64922751","wikidata":"https://www.wikidata.org/wiki/Q4650799","display_name":"Audio signal","level":3,"score":0.3050999939441681},{"id":"https://openalex.org/C104267543","wikidata":"https://www.wikidata.org/wiki/Q208163","display_name":"Signal processing","level":3,"score":0.2854999899864197},{"id":"https://openalex.org/C88639978","wikidata":"https://www.wikidata.org/wiki/Q233861","display_name":"Musical notation","level":3,"score":0.2838999927043915},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.28220000863075256},{"id":"https://openalex.org/C101468663","wikidata":"https://www.wikidata.org/wiki/Q1620158","display_name":"Modular design","level":2,"score":0.2759999930858612},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.27410000562667847},{"id":"https://openalex.org/C66024118","wikidata":"https://www.wikidata.org/wiki/Q1122506","display_name":"Computational model","level":2,"score":0.27160000801086426},{"id":"https://openalex.org/C129844170","wikidata":"https://www.wikidata.org/wiki/Q41299","display_name":"Quadratic equation","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.26829999685287476},{"id":"https://openalex.org/C77637269","wikidata":"https://www.wikidata.org/wiki/Q7002051","display_name":"Neural coding","level":2,"score":0.26409998536109924},{"id":"https://openalex.org/C66746571","wikidata":"https://www.wikidata.org/wiki/Q1134833","display_name":"ENCODE","level":3,"score":0.2606000006198883},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.26019999384880066},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2590999901294708},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.2556000053882599}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/apsipaasc65261.2025.11249088","is_oa":false,"landing_page_url":"https://doi.org/10.1109/apsipaasc65261.2025.11249088","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":9,"referenced_works":["https://openalex.org/W2906214917","https://openalex.org/W3209419355","https://openalex.org/W4224947526","https://openalex.org/W4225281045","https://openalex.org/W4385245566","https://openalex.org/W4388118708","https://openalex.org/W4388820924","https://openalex.org/W4400111311","https://openalex.org/W4406858688"],"related_works":[],"abstract_inverted_index":{"This":[0,162],"paper":[1],"investigates":[2],"automatic":[3],"piano":[4,34,185],"transcription":[5,155,186],"based":[6],"on":[7,131,170],"computationally-efficient":[8],"yet":[9],"high-performant":[10],"variants":[11],"of":[12,53,177],"the":[13,21,43,50,54,91,108,120,132,137,159,171,175],"Transformer":[14],"that":[15,101,136],"can":[16],"capture":[17],"longer-term":[18],"dependency":[19],"over":[20],"whole":[22,44],"musical":[23],"piece.":[24],"Recently,":[25],"transformer-based":[26],"sequence-to-sequence":[27],"models":[28],"have":[29],"demonstrated":[30],"excellent":[31],"performance":[32,156],"in":[33,64,68,144],"transcription.":[35],"These":[36],"models,":[37],"however,":[38],"fail":[39],"to":[40,49,103,107,124,158],"deal":[41],"with":[42,79,166],"piece":[45],"at":[46,192],"once":[47],"due":[48],"quadratic":[51],"complexity":[52],"self-attention":[55,87],"mechanism,":[56],"and":[57,93,95,122,147,183],"music":[58],"signals":[59],"are":[60],"thus":[61],"typically":[62],"processed":[63],"a":[65,96,115,141],"sliding-window":[66,86],"manner":[67],"practice.":[69],"To":[70],"overcome":[71],"this":[72],"limitation,":[73],"we":[74,84],"propose":[75],"an":[76],"efficient":[77,182],"architecture":[78],"sparse":[80,178],"attention":[81,179],"mechanisms.":[82],"Specifically,":[83],"introduce":[85],"mechanisms":[88],"for":[89,164,180],"both":[90],"encoder":[92,121],"decoder,":[94],"hybrid":[97],"global-local":[98],"cross-attention":[99],"mechanism":[100],"attends":[102],"various":[104],"spans":[105],"according":[106],"MIDI":[109],"token":[110],"types.":[111],"We":[112],"also":[113],"use":[114],"hierarchical":[116],"pooling":[117],"strategy":[118],"between":[119],"decoder":[123],"further":[125],"reduce":[126],"computational":[127,145],"load.":[128],"Our":[129],"experiments":[130],"MAESTRO":[133],"dataset":[134],"showed":[135],"proposed":[138],"model":[139],"achieved":[140],"significant":[142],"reduction":[143],"cost":[146],"memory":[148],"usage,":[149],"accelerating":[150],"inference":[151],"speed,":[152],"while":[153],"maintaining":[154],"comparable":[157],"full-attention":[160],"baseline.":[161],"allows":[163],"training":[165],"longer":[167],"audio":[168],"contexts":[169],"same":[172],"hardware,":[173],"demonstrating":[174],"viability":[176],"building":[181],"high-performance":[184],"systems.":[187],"The":[188],"code":[189],"is":[190],"available":[191],"https://github.com/WX-Wei/efficient-seq2seq-piano-trans.":[193]},"counts_by_year":[],"updated_date":"2026-06-26T08:34:08.712188","created_date":"2025-11-28T00:00:00"}