audio avatars
A²-LLM: End-to-End Audio Avatars via Large Language Models
New research presents A²-LLM, an end-to-end framework that unifies conversational AI with audio avatar generation, enabling seamless speech-driven digital humans through large language models.