IndicTTS

IndicTTS is a benchmark that evaluates text-to-speech synthesis models using a dataset containing high-quality speech data from multiple speakers across various Indian languages.

Papers using IndicTTS (4)

IndicVoices-R: Unlocking a Massive Multilingual Multi-speaker Speech Corpus for Scaling Indian TTS2024 · 1 citesProsody-TTS: An end-to-end speech synthesis system with prosody control2021DITTO: Data-efficient and Fair Targeted Subset Selection for ASR Accent Adaptation2021CrossVoice: Crosslingual Prosody Preserving Cascade-S2ST using Transfer Learning2024