OpenCodePapers

text-to-speech-synthesis-on-ljspeech

Text-To-Speech Synthesis
Dataset Link
Results over time
Click legend items to toggle metrics. Hover points for model names.
Leaderboard
Click a sortable column header to sort.
PaperCodeAudio Quality MOSPleasantness MOSWord Error Rate (WER)MOSWER (%)ModelNameReleaseDate
NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level Quality✓ Link4.56NaturalSpeech2022-05-09
NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level Quality✓ Link4.43VITS2022-05-09
Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech✓ Link4.37Grad-TTS + HiFiGAN (1000 steps)2021-05-13
Glow-TTS: A Generative Flow for Text-to-Speech via Monotonic Alignment Search✓ Link4.34Glow-TTS + HiFiGAN2020-05-22
NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level Quality✓ Link4.34FastSpeech 2 + HiFiGAN2022-05-09
FastSpeech 2: Fast and High-Quality End-to-End Text to Speech✓ Link4.32FastSpeech 2 + HiFiGAN2020-06-08
FastDiff: A Fast Conditional Diffusion Model for High-Quality Speech Synthesis✓ Link4.28FastDiff (4 steps)2022-04-21
FastDiff: A Fast Conditional Diffusion Model for High-Quality Speech Synthesis✓ Link4.03FastDiff-TTS2022-04-21
Neural Speech Synthesis with Transformer Network✓ Link3.88Transformer TTS (Mel + WaveGlow)2018-09-19
FastSpeech: Fast, Robust and Controllable Text to Speech✓ Link3.84FastSpeech (Mel + WaveGlow)2019-05-22
OverFlow: Putting flows on top of neural transducers for better TTS✓ Link3.372.30OverFlow2022-11-13
FastSpeech: Fast, Robust and Controllable Text to Speech✓ Link2.4Merlin2019-05-22
[]()1.25temp
Flowtron: an Autoregressive Flow-based Generative Network for Text-to-Speech Synthesis✓ Link3.665Flowtron2020-05-12
Flowtron: an Autoregressive Flow-based Generative Network for Text-to-Speech Synthesis✓ Link3.521Tacotron 22020-05-12
Matcha-TTS: A fast TTS architecture with conditional flow matching✓ Link3.842.09Matcha-TTS2023-09-06
Check out how to contribute new results.