HoliTok: A Continuous Holistic Tokenization with Robust Dual Capabilities of Speech Generation and Understanding
HoliTok is proposed, a continuous Holistic speech Tokenization model designed for unified generation-understanding modeling that achieves competitive reconstruction fidelity, improves generative learnability for high-quality and controllable synthesis, and is the only one that operates robustly in the authors' unified...
Bo-Han Li, Shiyue Lian, Hankun Wang et al.
· 3 citations