Zero-Shot Mono-to-Binaural Speech Synthesis

We present ZeroBAS, a neural method to synthesize binaural audio from monaural audio recordings and positional information without training on any binaural data. To our knowledge, this is the first published zero-shot neural approach to mono-to-binaural audio synthesis. Specifically, we show that a...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Levkovitch, Alon, Salazar, Julian, Mariooryad, Soroosh, Skerry-Ryan, RJ, Bar, Nadav, Kleijn, Bastiaan, Nachmani, Eliya
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Levkovitch, Alon
Salazar, Julian
Mariooryad, Soroosh
Skerry-Ryan, RJ
Bar, Nadav
Kleijn, Bastiaan
Nachmani, Eliya
description We present ZeroBAS, a neural method to synthesize binaural audio from monaural audio recordings and positional information without training on any binaural data. To our knowledge, this is the first published zero-shot neural approach to mono-to-binaural audio synthesis. Specifically, we show that a parameter-free geometric time warping and amplitude scaling based on source location suffices to get an initial binaural synthesis that can be refined by iteratively applying a pretrained denoising vocoder. Furthermore, we find this leads to generalization across room conditions, which we measure by introducing a new dataset, TUT Mono-to-Binaural, to evaluate state-of-the-art monaural-to-binaural synthesis methods on unseen conditions. Our zero-shot method is perceptually on-par with the performance of supervised methods on the standard mono-to-binaural dataset, and even surpasses them on our out-of-distribution TUT Mono-to-Binaural dataset. Our results highlight the potential of pretrained generative audio models and zero-shot learning to unlock robust binaural audio synthesis.
doi_str_mv 10.48550/arxiv.2412.08356
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2412_08356</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2412_08356</sourcerecordid><originalsourceid>FETCH-arxiv_primary_2412_083563</originalsourceid><addsrcrecordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjE00jOwMDY142TQjkotytcNzsgvUfDNz8vXLcnXdcrMSywtSsxRCC5ITU3OUAiuzCvJSC3OLOZhYE1LzClO5YXS3Azybq4hzh66YGPjC4oycxOLKuNBxseDjTcmrAIAFh8urQ</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Zero-Shot Mono-to-Binaural Speech Synthesis</title><source>arXiv.org</source><creator>Levkovitch, Alon ; Salazar, Julian ; Mariooryad, Soroosh ; Skerry-Ryan, RJ ; Bar, Nadav ; Kleijn, Bastiaan ; Nachmani, Eliya</creator><creatorcontrib>Levkovitch, Alon ; Salazar, Julian ; Mariooryad, Soroosh ; Skerry-Ryan, RJ ; Bar, Nadav ; Kleijn, Bastiaan ; Nachmani, Eliya</creatorcontrib><description>We present ZeroBAS, a neural method to synthesize binaural audio from monaural audio recordings and positional information without training on any binaural data. To our knowledge, this is the first published zero-shot neural approach to mono-to-binaural audio synthesis. Specifically, we show that a parameter-free geometric time warping and amplitude scaling based on source location suffices to get an initial binaural synthesis that can be refined by iteratively applying a pretrained denoising vocoder. Furthermore, we find this leads to generalization across room conditions, which we measure by introducing a new dataset, TUT Mono-to-Binaural, to evaluate state-of-the-art monaural-to-binaural synthesis methods on unseen conditions. Our zero-shot method is perceptually on-par with the performance of supervised methods on the standard mono-to-binaural dataset, and even surpasses them on our out-of-distribution TUT Mono-to-Binaural dataset. Our results highlight the potential of pretrained generative audio models and zero-shot learning to unlock robust binaural audio synthesis.</description><identifier>DOI: 10.48550/arxiv.2412.08356</identifier><language>eng</language><subject>Computer Science - Learning ; Computer Science - Sound</subject><creationdate>2024-12</creationdate><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2412.08356$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2412.08356$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Levkovitch, Alon</creatorcontrib><creatorcontrib>Salazar, Julian</creatorcontrib><creatorcontrib>Mariooryad, Soroosh</creatorcontrib><creatorcontrib>Skerry-Ryan, RJ</creatorcontrib><creatorcontrib>Bar, Nadav</creatorcontrib><creatorcontrib>Kleijn, Bastiaan</creatorcontrib><creatorcontrib>Nachmani, Eliya</creatorcontrib><title>Zero-Shot Mono-to-Binaural Speech Synthesis</title><description>We present ZeroBAS, a neural method to synthesize binaural audio from monaural audio recordings and positional information without training on any binaural data. To our knowledge, this is the first published zero-shot neural approach to mono-to-binaural audio synthesis. Specifically, we show that a parameter-free geometric time warping and amplitude scaling based on source location suffices to get an initial binaural synthesis that can be refined by iteratively applying a pretrained denoising vocoder. Furthermore, we find this leads to generalization across room conditions, which we measure by introducing a new dataset, TUT Mono-to-Binaural, to evaluate state-of-the-art monaural-to-binaural synthesis methods on unseen conditions. Our zero-shot method is perceptually on-par with the performance of supervised methods on the standard mono-to-binaural dataset, and even surpasses them on our out-of-distribution TUT Mono-to-Binaural dataset. Our results highlight the potential of pretrained generative audio models and zero-shot learning to unlock robust binaural audio synthesis.</description><subject>Computer Science - Learning</subject><subject>Computer Science - Sound</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjE00jOwMDY142TQjkotytcNzsgvUfDNz8vXLcnXdcrMSywtSsxRCC5ITU3OUAiuzCvJSC3OLOZhYE1LzClO5YXS3Azybq4hzh66YGPjC4oycxOLKuNBxseDjTcmrAIAFh8urQ</recordid><startdate>20241211</startdate><enddate>20241211</enddate><creator>Levkovitch, Alon</creator><creator>Salazar, Julian</creator><creator>Mariooryad, Soroosh</creator><creator>Skerry-Ryan, RJ</creator><creator>Bar, Nadav</creator><creator>Kleijn, Bastiaan</creator><creator>Nachmani, Eliya</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20241211</creationdate><title>Zero-Shot Mono-to-Binaural Speech Synthesis</title><author>Levkovitch, Alon ; Salazar, Julian ; Mariooryad, Soroosh ; Skerry-Ryan, RJ ; Bar, Nadav ; Kleijn, Bastiaan ; Nachmani, Eliya</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-arxiv_primary_2412_083563</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Learning</topic><topic>Computer Science - Sound</topic><toplevel>online_resources</toplevel><creatorcontrib>Levkovitch, Alon</creatorcontrib><creatorcontrib>Salazar, Julian</creatorcontrib><creatorcontrib>Mariooryad, Soroosh</creatorcontrib><creatorcontrib>Skerry-Ryan, RJ</creatorcontrib><creatorcontrib>Bar, Nadav</creatorcontrib><creatorcontrib>Kleijn, Bastiaan</creatorcontrib><creatorcontrib>Nachmani, Eliya</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Levkovitch, Alon</au><au>Salazar, Julian</au><au>Mariooryad, Soroosh</au><au>Skerry-Ryan, RJ</au><au>Bar, Nadav</au><au>Kleijn, Bastiaan</au><au>Nachmani, Eliya</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Zero-Shot Mono-to-Binaural Speech Synthesis</atitle><date>2024-12-11</date><risdate>2024</risdate><abstract>We present ZeroBAS, a neural method to synthesize binaural audio from monaural audio recordings and positional information without training on any binaural data. To our knowledge, this is the first published zero-shot neural approach to mono-to-binaural audio synthesis. Specifically, we show that a parameter-free geometric time warping and amplitude scaling based on source location suffices to get an initial binaural synthesis that can be refined by iteratively applying a pretrained denoising vocoder. Furthermore, we find this leads to generalization across room conditions, which we measure by introducing a new dataset, TUT Mono-to-Binaural, to evaluate state-of-the-art monaural-to-binaural synthesis methods on unseen conditions. Our zero-shot method is perceptually on-par with the performance of supervised methods on the standard mono-to-binaural dataset, and even surpasses them on our out-of-distribution TUT Mono-to-Binaural dataset. Our results highlight the potential of pretrained generative audio models and zero-shot learning to unlock robust binaural audio synthesis.</abstract><doi>10.48550/arxiv.2412.08356</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2412.08356
ispartof
issn
language eng
recordid cdi_arxiv_primary_2412_08356
source arXiv.org
subjects Computer Science - Learning
Computer Science - Sound
title Zero-Shot Mono-to-Binaural Speech Synthesis
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-06T14%3A46%3A02IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Zero-Shot%20Mono-to-Binaural%20Speech%20Synthesis&rft.au=Levkovitch,%20Alon&rft.date=2024-12-11&rft_id=info:doi/10.48550/arxiv.2412.08356&rft_dat=%3Carxiv_GOX%3E2412_08356%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true