1 paper across 1 session
We propose a proactive defensive framework against malicious production LLM-based speech synthesis to protect our voice information.
