{"id":"https://openalex.org/W4416251049","doi":"https://doi.org/10.1109/waspaa66052.2025.11230962","title":"Post-Training Quantization for Audio Diffusion Transformers","display_name":"Post-Training Quantization for Audio Diffusion Transformers","publication_year":2025,"publication_date":"2025-10-12","ids":{"openalex":"https://openalex.org/W4416251049","doi":"https://doi.org/10.1109/waspaa66052.2025.11230962"},"language":null,"primary_location":{"id":"doi:10.1109/waspaa66052.2025.11230962","is_oa":false,"landing_page_url":"https://doi.org/10.1109/waspaa66052.2025.11230962","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)","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/A5099181209","display_name":"Tanmay Khandelwal","orcid":"https://orcid.org/0009-0004-3770-8317"},"institutions":[{"id":"https://openalex.org/I36672615","display_name":"Courant Institute of Mathematical Sciences","ror":"https://ror.org/037tm7f56","country_code":"US","type":"education","lineage":["https://openalex.org/I36672615","https://openalex.org/I57206974"]},{"id":"https://openalex.org/I57206974","display_name":"New York University","ror":"https://ror.org/0190ak572","country_code":"US","type":"education","lineage":["https://openalex.org/I57206974"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Tanmay Khandelwal","raw_affiliation_strings":["New York University,Courant Institute of Mathematical Sciences,NY,USA"],"affiliations":[{"raw_affiliation_string":"New York University,Courant Institute of Mathematical Sciences,NY,USA","institution_ids":["https://openalex.org/I36672615","https://openalex.org/I57206974"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5021235229","display_name":"Magdalena Fuentes","orcid":"https://orcid.org/0000-0003-4506-6639"},"institutions":[{"id":"https://openalex.org/I57206974","display_name":"New York University","ror":"https://ror.org/0190ak572","country_code":"US","type":"education","lineage":["https://openalex.org/I57206974"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Magdalena Fuentes","raw_affiliation_strings":["New York University,MARL,NY,USA"],"affiliations":[{"raw_affiliation_string":"New York University,MARL,NY,USA","institution_ids":["https://openalex.org/I57206974"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5099181209"],"corresponding_institution_ids":["https://openalex.org/I36672615","https://openalex.org/I57206974"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.37390674,"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":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11349","display_name":"Music Technology and Sound Studies","score":0.2612000107765198,"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"}},"topics":[{"id":"https://openalex.org/T11349","display_name":"Music Technology and Sound Studies","score":0.2612000107765198,"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/T10283","display_name":"Hearing Loss and Rehabilitation","score":0.12870000302791595,"subfield":{"id":"https://openalex.org/subfields/2805","display_name":"Cognitive Neuroscience"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.09049999713897705,"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/quantization","display_name":"Quantization (signal processing)","score":0.7402999997138977},{"id":"https://openalex.org/keywords/smoothing","display_name":"Smoothing","score":0.5026999711990356},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.49570000171661377},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.4535999894142151},{"id":"https://openalex.org/keywords/high-fidelity","display_name":"High fidelity","score":0.44530001282691956},{"id":"https://openalex.org/keywords/vector-quantization","display_name":"Vector quantization","score":0.4406999945640564},{"id":"https://openalex.org/keywords/fidelity","display_name":"Fidelity","score":0.42570000886917114},{"id":"https://openalex.org/keywords/speech-coding","display_name":"Speech coding","score":0.37700000405311584}],"concepts":[{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.7402999997138977},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6033999919891357},{"id":"https://openalex.org/C3770464","wikidata":"https://www.wikidata.org/wiki/Q775963","display_name":"Smoothing","level":2,"score":0.5026999711990356},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.49570000171661377},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.48899999260902405},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.4535999894142151},{"id":"https://openalex.org/C113364801","wikidata":"https://www.wikidata.org/wiki/Q26674","display_name":"High fidelity","level":2,"score":0.44530001282691956},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.4406999945640564},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.42570000886917114},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.4169999957084656},{"id":"https://openalex.org/C13895895","wikidata":"https://www.wikidata.org/wiki/Q3270773","display_name":"Speech coding","level":2,"score":0.37700000405311584},{"id":"https://openalex.org/C24326235","wikidata":"https://www.wikidata.org/wiki/Q126095","display_name":"Electronic engineering","level":1,"score":0.3662000000476837},{"id":"https://openalex.org/C188198153","wikidata":"https://www.wikidata.org/wiki/Q1613840","display_name":"Limiting","level":2,"score":0.3610000014305115},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.305400013923645},{"id":"https://openalex.org/C22789450","wikidata":"https://www.wikidata.org/wiki/Q420904","display_name":"Singular value decomposition","level":2,"score":0.30140000581741333},{"id":"https://openalex.org/C87133666","wikidata":"https://www.wikidata.org/wiki/Q1161699","display_name":"Dynamic range","level":2,"score":0.30079999566078186},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2858999967575073},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.2858999967575073},{"id":"https://openalex.org/C2780056265","wikidata":"https://www.wikidata.org/wiki/Q106239881","display_name":"High dynamic range","level":3,"score":0.28209999203681946},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.27900001406669617},{"id":"https://openalex.org/C2780388253","wikidata":"https://www.wikidata.org/wiki/Q5421508","display_name":"Exponent","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C29265498","wikidata":"https://www.wikidata.org/wiki/Q7047719","display_name":"Noise measurement","level":3,"score":0.2565000057220459}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/waspaa66052.2025.11230962","is_oa":false,"landing_page_url":"https://doi.org/10.1109/waspaa66052.2025.11230962","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)","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":10,"referenced_works":["https://openalex.org/W2945761034","https://openalex.org/W4366392438","https://openalex.org/W4378902066","https://openalex.org/W4386047807","https://openalex.org/W4386065704","https://openalex.org/W4390872297","https://openalex.org/W4401980590","https://openalex.org/W4402753724","https://openalex.org/W4408352114","https://openalex.org/W4409262168"],"related_works":[],"abstract_inverted_index":{"Diffusion":[0],"Transformers":[1],"(DiTs)":[2],"enable":[3],"high-quality":[4],"audio":[5,35],"synthesis":[6],"but":[7],"are":[8],"often":[9],"computationally":[10],"intensive":[11],"and":[12,42,61,67,94,100],"require":[13],"substantial":[14],"storage,":[15],"which":[16],"limits":[17],"their":[18],"practical":[19,49],"deployment.":[20],"In":[21],"this":[22],"paper,":[23],"we":[24,91],"present":[25],"a":[26,52,69],"comprehensive":[27],"evaluation":[28],"of":[29],"post-training":[30],"quantization":[31,44,58,109],"(PTQ)":[32],"techniques":[33],"for":[34,83],"DiTs,":[36],"analyzing":[37],"the":[38],"trade-offs":[39],"between":[40],"static":[41,117],"dynamic":[43,108],"schemes.":[45],"We":[46],"explore":[47],"two":[48],"extensions":[50],"(1)":[51],"denoising-timestep-aware":[53],"smoothing":[54],"method":[55],"that":[56,107,128],"adapts":[57],"scales":[59],"perinput-channel":[60],"timestep":[62],"to":[63,81,141],"mitigate":[64],"activation":[65],"outliers,":[66],"(2)":[68],"lightweight":[70],"low-rank":[71],"adapter":[72],"(LoRA)-based":[73],"branch":[74],"derived":[75],"from":[76],"singular":[77],"value":[78],"decomposition":[79],"(SVD)":[80],"compensate":[82],"residual":[84],"weight":[85],"errors.":[86],"Using":[87],"Stable":[88],"Audio":[89],"Open":[90],"benchmark":[92],"W8A8":[93],"W4A8":[95],"configurations":[96],"across":[97],"objective":[98],"metrics":[99],"human":[101],"perceptual":[102],"ratings.":[103],"Our":[104],"results":[105],"show":[106,127],"preserves":[110],"fidelity":[111],"even":[112],"at":[113],"lower":[114,122],"precision,":[115],"while":[116,135],"methods":[118],"remain":[119],"competitive":[120],"with":[121],"latency.":[123],"Overall,":[124],"our":[125],"findings":[126],"low-precision":[129],"DiTs":[130],"can":[131],"retain":[132],"high-fidelity":[133],"generation":[134],"reducing":[136],"memory":[137],"usage":[138],"by":[139],"up":[140],"79%.":[142]},"counts_by_year":[],"updated_date":"2026-03-07T16:01:11.037858","created_date":"2025-11-14T00:00:00"}