Voice Attacker: Multi-Head Factorized Attentive Reconstructor and Gradient Reversal for Random Prosody Anonymization
Published in Voice Privacy 2024, 2024
Authors: Do Tri Nhan et al. (2024). Voice Attacker: Multi-Head Factorized Attentive Reconstructor and Gradient Reversal for Random Prosody Anonymization. Voice Privacy Challenge 2024.
A speaker anonymization system using multi-head factorized attentive reconstruction and gradient reversal layers to achieve random prosody anonymization while preserving linguistic content.