Speech privacy-preserving methods using secret key for convolutional neural network models and their robustness evaluation

In this paper, we propose privacy-preserving methods with a secret key for convolutional neural network (CNN)-based models in speech processing tasks. In environments where untrusted third parties, like cloud servers, provide CNN-based systems, ensuring the privacy of speech queries becomes essentia...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Niwa, Shoko, Shiota, Sayaka, Kiya, Hitoshi
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 Niwa, Shoko
Shiota, Sayaka
Kiya, Hitoshi
description In this paper, we propose privacy-preserving methods with a secret key for convolutional neural network (CNN)-based models in speech processing tasks. In environments where untrusted third parties, like cloud servers, provide CNN-based systems, ensuring the privacy of speech queries becomes essential. This paper proposes encryption methods for speech queries using secret keys and a model structure that allows for encrypted queries to be accepted without decryption. Our approach introduces three types of secret keys: Shuffling, Flipping, and random orthogonal matrix (ROM). In experiments, we demonstrate that when the proposed methods are used with the correct key, identification performance did not degrade. Conversely, when an incorrect key is used, the performance significantly decreased. Particularly, with the use of ROM, we show that even with a relatively small key space, high privacy-preserving performance can be maintained many speech processing tasks. Furthermore, we also demonstrate the difficulty of recovering original speech from encrypted queries in various robustness evaluations.
doi_str_mv 10.48550/arxiv.2408.03897
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2408_03897</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2408_03897</sourcerecordid><originalsourceid>FETCH-arxiv_primary_2408_038973</originalsourceid><addsrcrecordid>eNqFzsEOgjAQBNBePBj1Azy5PyCiYMSz0XjXO6mwSGNpyW5bxa9XiHdPk0lmkifEfB1HabbdxitJLxWiTRpnUZxk-91YvC8tYlFDSyrIolu2hIwUlLlDg662JYPnvjEWhA4e2EFlCQprgtXeKWukBoOehnBPSw9obImaQZoSXI2KgOzNszPIDBik9rL_TcWokppx9suJWJyO18N5OSjzr6iR1OW9Nh-0yf_FB5JDTmI</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Speech privacy-preserving methods using secret key for convolutional neural network models and their robustness evaluation</title><source>arXiv.org</source><creator>Niwa, Shoko ; Shiota, Sayaka ; Kiya, Hitoshi</creator><creatorcontrib>Niwa, Shoko ; Shiota, Sayaka ; Kiya, Hitoshi</creatorcontrib><description>In this paper, we propose privacy-preserving methods with a secret key for convolutional neural network (CNN)-based models in speech processing tasks. In environments where untrusted third parties, like cloud servers, provide CNN-based systems, ensuring the privacy of speech queries becomes essential. This paper proposes encryption methods for speech queries using secret keys and a model structure that allows for encrypted queries to be accepted without decryption. Our approach introduces three types of secret keys: Shuffling, Flipping, and random orthogonal matrix (ROM). In experiments, we demonstrate that when the proposed methods are used with the correct key, identification performance did not degrade. Conversely, when an incorrect key is used, the performance significantly decreased. Particularly, with the use of ROM, we show that even with a relatively small key space, high privacy-preserving performance can be maintained many speech processing tasks. Furthermore, we also demonstrate the difficulty of recovering original speech from encrypted queries in various robustness evaluations.</description><identifier>DOI: 10.48550/arxiv.2408.03897</identifier><language>eng</language><subject>Computer Science - Cryptography and Security ; Computer Science - Sound</subject><creationdate>2024-08</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/2408.03897$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2408.03897$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Niwa, Shoko</creatorcontrib><creatorcontrib>Shiota, Sayaka</creatorcontrib><creatorcontrib>Kiya, Hitoshi</creatorcontrib><title>Speech privacy-preserving methods using secret key for convolutional neural network models and their robustness evaluation</title><description>In this paper, we propose privacy-preserving methods with a secret key for convolutional neural network (CNN)-based models in speech processing tasks. In environments where untrusted third parties, like cloud servers, provide CNN-based systems, ensuring the privacy of speech queries becomes essential. This paper proposes encryption methods for speech queries using secret keys and a model structure that allows for encrypted queries to be accepted without decryption. Our approach introduces three types of secret keys: Shuffling, Flipping, and random orthogonal matrix (ROM). In experiments, we demonstrate that when the proposed methods are used with the correct key, identification performance did not degrade. Conversely, when an incorrect key is used, the performance significantly decreased. Particularly, with the use of ROM, we show that even with a relatively small key space, high privacy-preserving performance can be maintained many speech processing tasks. Furthermore, we also demonstrate the difficulty of recovering original speech from encrypted queries in various robustness evaluations.</description><subject>Computer Science - Cryptography and Security</subject><subject>Computer Science - Sound</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNqFzsEOgjAQBNBePBj1Azy5PyCiYMSz0XjXO6mwSGNpyW5bxa9XiHdPk0lmkifEfB1HabbdxitJLxWiTRpnUZxk-91YvC8tYlFDSyrIolu2hIwUlLlDg662JYPnvjEWhA4e2EFlCQprgtXeKWukBoOehnBPSw9obImaQZoSXI2KgOzNszPIDBik9rL_TcWokppx9suJWJyO18N5OSjzr6iR1OW9Nh-0yf_FB5JDTmI</recordid><startdate>20240807</startdate><enddate>20240807</enddate><creator>Niwa, Shoko</creator><creator>Shiota, Sayaka</creator><creator>Kiya, Hitoshi</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20240807</creationdate><title>Speech privacy-preserving methods using secret key for convolutional neural network models and their robustness evaluation</title><author>Niwa, Shoko ; Shiota, Sayaka ; Kiya, Hitoshi</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-arxiv_primary_2408_038973</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Cryptography and Security</topic><topic>Computer Science - Sound</topic><toplevel>online_resources</toplevel><creatorcontrib>Niwa, Shoko</creatorcontrib><creatorcontrib>Shiota, Sayaka</creatorcontrib><creatorcontrib>Kiya, Hitoshi</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Niwa, Shoko</au><au>Shiota, Sayaka</au><au>Kiya, Hitoshi</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Speech privacy-preserving methods using secret key for convolutional neural network models and their robustness evaluation</atitle><date>2024-08-07</date><risdate>2024</risdate><abstract>In this paper, we propose privacy-preserving methods with a secret key for convolutional neural network (CNN)-based models in speech processing tasks. In environments where untrusted third parties, like cloud servers, provide CNN-based systems, ensuring the privacy of speech queries becomes essential. This paper proposes encryption methods for speech queries using secret keys and a model structure that allows for encrypted queries to be accepted without decryption. Our approach introduces three types of secret keys: Shuffling, Flipping, and random orthogonal matrix (ROM). In experiments, we demonstrate that when the proposed methods are used with the correct key, identification performance did not degrade. Conversely, when an incorrect key is used, the performance significantly decreased. Particularly, with the use of ROM, we show that even with a relatively small key space, high privacy-preserving performance can be maintained many speech processing tasks. Furthermore, we also demonstrate the difficulty of recovering original speech from encrypted queries in various robustness evaluations.</abstract><doi>10.48550/arxiv.2408.03897</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2408.03897
ispartof
issn
language eng
recordid cdi_arxiv_primary_2408_03897
source arXiv.org
subjects Computer Science - Cryptography and Security
Computer Science - Sound
title Speech privacy-preserving methods using secret key for convolutional neural network models and their robustness evaluation
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-25T09%3A36%3A25IST&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=Speech%20privacy-preserving%20methods%20using%20secret%20key%20for%20convolutional%20neural%20network%20models%20and%20their%20robustness%20evaluation&rft.au=Niwa,%20Shoko&rft.date=2024-08-07&rft_id=info:doi/10.48550/arxiv.2408.03897&rft_dat=%3Carxiv_GOX%3E2408_03897%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