Framework for removing speaker voices from AI models without forgetting prior removals

Original: Continual Speaker Identity Unlearning with Minimal Interference

Source: arxiv.org ↗