Evaluating Voice Conversion-based Privacy Protection against Informed Attackers
This addresses privacy risks in speech data for applications like voice assistants, but it is incremental as it builds on prior anonymization work by evaluating specific attack scenarios.
The paper investigated voice conversion methods for anonymizing speech to protect speaker identity, finding that they fail against attackers with extensive knowledge of the conversion scheme but offer some protection against less informed attackers, with privacy measured by increases in equal error rate up to 10% in certain scenarios.
Speech data conveys sensitive speaker attributes like identity or accent. With a small amount of found data, such attributes can be inferred and exploited for malicious purposes: voice cloning, spoofing, etc. Anonymization aims to make the data unlinkable, i.e., ensure that no utterance can be linked to its original speaker. In this paper, we investigate anonymization methods based on voice conversion. In contrast to prior work, we argue that various linkage attacks can be designed depending on the attackers' knowledge about the anonymization scheme. We compare two frequency warping-based conversion methods and a deep learning based method in three attack scenarios. The utility of converted speech is measured via the word error rate achieved by automatic speech recognition, while privacy protection is assessed by the increase in equal error rate achieved by state-of-the-art i-vector or x-vector based speaker verification. Our results show that voice conversion schemes are unable to effectively protect against an attacker that has extensive knowledge of the type of conversion and how it has been applied, but may provide some protection against less knowledgeable attackers.