Pairing Weak with Strong: Twin Models for Defending Against Adversarial Attack on Speaker Verification.

Zhiyuan Peng, Xu Li, Tan Lee

Browse the full Interspeech paper archive.