# Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions

This paper describes Tacotron 2, a neural network architecture for speech synthesis directly from text. The system is composed of a recurrent sequence-to-sequence feature prediction network that maps character embeddings to mel-scale spectrograms, followed by a modified WaveNet model acting as a vocoder to synthesize timedomain waveforms from those spectrograms. Our model achieves a mean opinion score (MOS) of $4.53$ comparable to a MOS of $4.58$ for professionally recorded speech. To validate our design choices, we present ablation studies of key components of our system and evaluate the impact of using mel spectrograms as the input to WaveNet instead of linguistic, duration, and $F_0$ features. We further demonstrate that using a compact acoustic intermediate representation enables significant simplification of the WaveNet architecture.

PDF Abstract

## Code Add Remove Mark official

4,783
3,575
 ↳ Quickstart in
3,053
 ↳ Quickstart in
2,669
See all 26 implementations

## Datasets

Add Datasets introduced or used in this paper

## Results from the Paper Edit

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Speech Synthesis North American English Tacotron 2 Mean Opinion Score 4.526 # 1
Speech Synthesis North American English WaveNet (Linguistic) Mean Opinion Score 4.341 # 2