To cope with this problem, we propose a new speech-driven 3D facial animation framework RQTalker based on the Region-aware Vector Quantization mechanism. The key insight is to first build a ...