Speaker de-identification using diphone recognition and speech synthesis

Justin, Tadej; Dobrišek, Simon; Vesnicer, Boštjan; Ipšić, Ivo; Mihelic, F.

doi:10.1109/fg.2015.7285021

Cited by 33 publications

(21 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [156], an approach to voice de-identification based on a combination of diphone recognition and speech synthesis is proposed. De-identification is performed in two steps.…”

Section: Voice De-identificationmentioning

confidence: 99%

De-identification for privacy protection in multimedia content: A survey

Ribarić

Ariyaeeinia

Pavešić

2016

Signal Processing: Image Communication

156

107

View full text Add to dashboard Cite

a b s t r a c tPrivacy is one of the most important social and political issues in our information society, characterized by a growing range of enabling and supporting technologies and services. Amongst these are communications, multimedia, biometrics, big data, cloud computing, data mining, internet, social networks, and audio-video surveillance. Each of these can potentially provide the means for privacy intrusion. Deidentification is one of the main approaches to privacy protection in multimedia contents (text, still images, audio and video sequences and their combinations). It is a process for concealing or removing personal identifiers, or replacing them by surrogate personal identifiers in personal information in order to prevent the disclosure and use of data for purposes unrelated to the purpose for which the information was originally obtained. Based on the proposed taxonomy inspired by the Safe Harbour approach, the personal identifiers, i.e., the personal identifiable information, are classified as non-biometric, physiological and behavioural biometric, and soft biometric identifiers. In order to protect the privacy of an individual, all of the above identifiers will have to be de-identified in multimedia content. This paper presents a review of the concepts of privacy and the linkage among privacy, privacy protection, and the methods and technologies designed specifically for privacy protection in multimedia contents. The study provides an overview of de-identification approaches for non-biometric identifiers (text, hairstyle, dressing style, license plates), as well as for the physiological (face, fingerprint, iris, ear), behavioural (voice, gait, gesture) and soft-biometric (body silhouette, gender, age, race, tattoo) identifiers in multimedia documents.

show abstract

“…In [156], an approach to voice de-identification based on a combination of diphone recognition and speech synthesis is proposed. De-identification is performed in two steps.…”

Section: Voice De-identificationmentioning

confidence: 99%

De-identification for privacy protection in multimedia content: A survey

Ribarić

Ariyaeeinia

Pavešić

2016

Signal Processing: Image Communication

156

107

View full text Add to dashboard Cite

show abstract

“…Biometric privacy enhancing techniques (B-PETs) are in general applicable to different biometric characteristics, as evidenced by the vast body of research on this topic e.g., [32], [86], [87], [88], [89], [90], [91], [92], [93], [94], [95]. However, a considerable portion of the existing work is focusing on facial biometrics, mainly because of the multitude of (sensitive) information that can be inferred automatically by analyzing facial appearances and the fact that the analysis can be done without the cooperation and consent of individuals.…”

Section: Survey Of B-pets For Face Biometricsmentioning

confidence: 99%

Privacy–Enhancing Face Biometrics: A Comprehensive Survey

Meden

Rot

Terhörst

et al. 2021

IEEE Trans.Inform.Forensic Secur.

Self Cite

126

View full text Add to dashboard Cite

Biometric recognition technology has made significant advances over the last decade and is now used across a number of services and applications. However, this widespread deployment has also resulted in privacy concerns and evolving societal expectations about the appropriate use of the technology. For example, the ability to automatically extract age, gender, race, and health cues from biometric data has heightened concerns about privacy leakage. Face recognition technology, in particular, has been in the spotlight, and is now seen by many as posing a considerable risk to personal privacy. In response to these and similar concerns, researchers have intensified efforts towards developing techniques and computational models capable of ensuring privacy to individuals, while still facilitating the utility of face recognition technology in several application scenarios. These efforts have resulted in a multitude of privacy-enhancing techniques that aim at addressing privacy risks originating from biometric systems and providing technological solutions for legislative requirements set forth in privacy laws and regulations, such as GDPR. The goal of this overview paper is to provide a comprehensive introduction into privacy-related research in the area of biometrics and review existing work on Biometric Privacy-Enhancing Techniques (B-PETs) applied to face biometrics. To make this work useful for as wide of an audience as possible, several key topics are covered as well, including evaluation strategies used with B-PETs, existing datasets, relevant standards, and regulations and critical open issues that will have to be addressed in the future.

show abstract

“…Speaker de-identification is the task of concealing speaker identity, which may be revealed in linguistic (content of speaker's speech) [1,2] and paralinguistic (spectral and excitation features of the speech signal uttered by the speaker) [3,4,5] features. In this study, we focus on the latter one.…”

Section: Introductionmentioning

confidence: 99%

Convolutional Neural Network Based Speaker De-Identification

Bahmaninezhad¹,

Zhang²,

Hansen³

2018

The Speaker and Language Recognition Workshop (Odyssey 2018)

View full text Add to dashboard Cite

Concealing speaker identity in speech signals refers to the task of speaker de-identification, which helps protect the privacy of a speaker. Although, both linguistic and paralinguistic features reveal personal information of a speaker and they both need to be addressed, in this study we only focus on speaker voice characteristics. In other words, our goal is to move away from the source speaker identity while preserving naturalness and quality. The proposed speaker de-identification system maps voice of a given speaker to an average (or gender-dependent average) voice; the mapping is modeled by a new convolutional neural network (CNN) encoder-decoder architecture. Here, the transformation of both spectral and excitation features are studied. The voice conversion challenge 2016 (VCC-2016) database is used to train the system and examine performance of the proposed method. We use two different approaches for evaluations:(1) objective evaluation: equal error rates (EERs) calculated by an i-vector/PLDA speaker recognition system range between 1.265 -3.46 % on average for all developed systems, and (2) subjective evaluation: achieved 2.8 naturalness mean opinion score (MOS). Both objective and subjective experiments confirm the effectiveness of our proposed de-identification method.

show abstract

Speaker de-identification using diphone recognition and speech synthesis

Cited by 33 publications

References 9 publications

De-identification for privacy protection in multimedia content: A survey

De-identification for privacy protection in multimedia content: A survey

Privacy–Enhancing Face Biometrics: A Comprehensive Survey

Convolutional Neural Network Based Speaker De-Identification

Contact Info

Product

Resources

About