Defending Your Voice: Adversarial Attack on Voice Conversion
Authors:
- 19 January 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 552-559
Abstract
Substantial improvements have been achieved in recent years in voice conversion, which converts the speaker characteristics of an utterance into those of another speaker without changing the linguistic content of the utterance. Nonetheless, the improved conversion technologies also led to concerns about privacy and authentication. It thus becomes highly desired to be able to prevent one’s voice from being improperly utilized with such voice conversion technologies. This is why we report in this paper the first known attempt to perform adversarial attack on voice conversion. We introduce human imperceptible noise into the utterances of a speaker whose voice is to be defended. Given these adversarial examples, voice conversion models cannot convert other utterances so as to sound like being produced by the defended speaker. Preliminary experiments were conducted on two currently state-of-the-art zero-shot voice conversion models. Objective and subjective evaluation results in both white-box and black-box scenarios are reported. It was shown that the speaker characteristics of the converted utterances were made obviously different from those of the defended speaker, while the adversarial examples of the defended speaker are not distinguishable from the authentic utterances.Keywords
All Related Versions
This publication has 25 references indexed in Scilit:
- An end-to-end spoofing countermeasure for automatic speaker verification using evolving recurrent neural networks2018
- Audio Adversarial Examples: Targeted Attacks on Speech-to-Text2018
- Adversarial Examples for Generative Models2018
- VoxCeleb: A Large-Scale Speaker Identification Dataset2017
- Audio Replay Attack Detection with Deep Learning Frameworks2017
- You Can Hear But You Cannot Steal: Defending Against Voice Impersonation Attacks on Smartphones2017
- Towards Evaluating the Robustness of Neural Networks2017
- Deep features for automatic spoofing detectionSpeech Communication, 2016
- Librispeech: An ASR corpus based on public domain audio books2015
- Signal estimation from modified short-time Fourier transformIEEE Transactions on Acoustics, Speech, and Signal Processing, 1984