Mitigating Unauthorized Speech Synthesis for Voice Protection