VoxMel TTS: tokenizer-free autoregressive speech synthesis via Haar-factorized mel

VoxDance-Lab · hf · 2026-11-17

VoxDance Lab released the VoxMel TTS technical report, a tokenizer-free autoregressive text-to-speech framework.

Skipping discrete acoustic tokens avoids their information loss — an interesting architecture direction for high-fidelity TTS.

Original post →

More from Multimodal

Multimodal channel →