1 paper
Wataru Nakata, Yuki Saito, Yota Ueda +1
Large-scale text-to-speech (TTS) systems are limited by the scarcity of clean, multilingual recordings. We introduce Sidon, a fast, open-source speech restoration model that conver…