Adversarial Disentanglement of Speaker Representation for Attribute-Driven Privacy Preservation
This work addresses the privacy concern of sensitive personal attributes embedded in speaker voice representations for individuals using speech technologies, offering a method to conceal specific information.
This paper introduces attribute-driven privacy preservation for speaker voice representations, allowing users to hide specific personal attributes. They propose an adversarial autoencoding method to disentangle and conceal the sex attribute, demonstrating its effectiveness on VoxCeleb datasets while maintaining Automatic Speaker Verification (ASV) ability.
In speech technologies, speaker's voice representation is used in many applications such as speech recognition, voice conversion, speech synthesis and, obviously, user authentication. Modern vocal representations of the speaker are based on neural embeddings. In addition to the targeted information, these representations usually contain sensitive information about the speaker, like the age, sex, physical state, education level or ethnicity. In order to allow the user to choose which information to protect, we introduce in this paper the concept of attribute-driven privacy preservation in speaker voice representation. It allows a person to hide one or more personal aspects to a potential malicious interceptor and to the application provider. As a first solution to this concept, we propose to use an adversarial autoencoding method that disentangles in the voice representation a given speaker attribute thus allowing its concealment. We focus here on the sex attribute for an Automatic Speaker Verification (ASV) task. Experiments carried out using the VoxCeleb datasets have shown that the proposed method enables the concealment of this attribute while preserving ASV ability.