keonlee9420/StyleSpeech keonlee9420/StyleSpeech

  • Stars
    star
    177
  • Rank 208,134 (Top 5 %)
  • Language
    Python
  • License
    MIT License
  • Created almost 3 years ago
  • Updated about 2 years ago
Twitter logo Share LinkedIn logo Share Facebook logo Share
keonlee9420/StyleSpeech
github

keonlee9420

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

PyTorch Implementation of Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation

StyleSpeech - PyTorch Implementation

PyTorch Implementation of Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation.

Branch

  • StyleSpeech (naive branch)
  • Meta-StyleSpeech (main branch)

Quickstart

Dependencies

You can install the Python dependencies with

pip3 install -r requirements.txt

Inference

You have to download pretrained models and put them in output/ckpt/LibriTTS_meta_learner/.

For English multi-speaker TTS, run

python3 synthesize.py --text "YOUR_DESIRED_TEXT" --ref_audio path/to/reference_audio.wav --restore_step 200000 --mode single -p config/LibriTTS/preprocess.yaml -m config/LibriTTS/model.yaml -t config/LibriTTS/train.yaml

The generated utterances will be put in output/result/. Your synthesized speech will have ref_audio's style.

Batch Inference

Batch inference is also supported, try

python3 synthesize.py --source preprocessed_data/LibriTTS/val.txt --restore_step 200000 --mode batch -p config/LibriTTS/preprocess.yaml -m config/LibriTTS/model.yaml -t config/LibriTTS/train.yaml

to synthesize all utterances in preprocessed_data/LibriTTS/val.txt. This can be viewed as a reconstruction of validation datasets referring to themselves for the reference style.

Controllability

The pitch/volume/speaking rate of the synthesized utterances can be controlled by specifying the desired pitch/energy/duration ratios. For example, one can increase the speaking rate by 20 % and decrease the volume by 20 % by

python3 synthesize.py --text "YOUR_DESIRED_TEXT" --restore_step 200000 --mode single -p config/LibriTTS/preprocess.yaml -m config/LibriTTS/model.yaml -t config/LibriTTS/train.yaml --duration_control 0.8 --energy_control 0.8

Note that the controllability is originated from FastSpeech2 and not a vital interest of StyleSpeech. Please refer to STYLER [demo, code] for the controllability of each style factor.

Training

Datasets

The supported datasets are

  • LibriTTS: a multi-speaker English dataset containing 585 hours of speech by 2456 speakers.
  • (will be added more)

Preprocessing

Run

python3 prepare_align.py config/LibriTTS/preprocess.yaml

for some preparations.

For the forced alignment, Montreal Forced Aligner (MFA) is used to obtain the alignments between the utterances and the phoneme sequences. Pre-extracted alignments for the datasets are provided here. You have to unzip the files in preprocessed_data/LibriTTS/TextGrid/. Alternately, you can run the aligner by yourself.

After that, run the preprocessing script by

python3 preprocess.py config/LibriTTS/preprocess.yaml

Training

Train your model with

python3 train.py -p config/LibriTTS/preprocess.yaml -m config/LibriTTS/model.yaml -t config/LibriTTS/train.yaml

As described in the paper, the script will start from pre-training the naive model until meta_learning_warmup steps and then meta-train the model for additional steps via episodic training.

TensorBoard

Use

tensorboard --logdir output/log/LibriTTS

to serve TensorBoard on your localhost. The loss curves, synthesized mel-spectrograms, and audios are shown.

Implementation Issues

  1. Use 22050Hz sampling rate instead of 16kHz.
  2. Add one fully connected layer at the beginning of Mel-Style Encoder to upsample input mel-spectrogram from 80 to 128.
  3. The model size including meta-learner is 28.197M.
  4. Use a maximum 16 batch size on training instead of 48 or 20 mainly due to the lack of memory capacity with a single 24GiB TITAN-RTX. This can be achieved by the following script to filter out data longer than max_seq_len: 
    python3 filelist_filtering.py -p config/LibriTTS/preprocess.yaml -m config/LibriTTS/model.yaml
    This will generate train_filtered.txt in the same location of train.txt.
  5. Since the total batch size is decreased, the number of training steps is doubled compared to the original paper.
  6. Use HiFi-GAN instead of MelGAN for vocoding.

Citation

@misc{lee2021stylespeech,
  author = {Lee, Keon},
  title = {StyleSpeech},
  year = {2021},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/keonlee9420/StyleSpeech}}
}

References

More Repositories

1

PortaSpeech

PyTorch Implementation of PortaSpeech: Portable and High-Quality Generative Text-to-Speech
Python
330
star
2

Comprehensive-Transformer-TTS

A Non-Autoregressive Transformer based Text-to-Speech, supporting a family of SOTA transformers with supervised and unsupervised duration modelings. This project grows with the research community, aiming to achieve the ultimate TTS
Python
308
star
3

DiffGAN-TTS

PyTorch Implementation of DiffGAN-TTS: High-Fidelity and Efficient Text-to-Speech with Denoising Diffusion GANs
Python
293
star
4

Expressive-FastSpeech2

PyTorch Implementation of Non-autoregressive Expressive (emotional, conversational) TTS based on FastSpeech2, supporting English, Korean, and your own languages.
Python
256
star
5

DiffSinger

PyTorch implementation of DiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism (focused on DiffSpeech)
Python
220
star
6

Parallel-Tacotron2

PyTorch Implementation of Google's Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling
Python
186
star
7

DailyTalk

Official repository of DailyTalk: Spoken Dialogue Dataset for Conversational Text-to-Speech, ICASSP 2023 (Oral)
Python
175
star
8

Cross-Speaker-Emotion-Transfer

PyTorch Implementation of ByteDance's Cross-speaker Emotion Transfer Based on Speaker Condition Layer Normalization and Semi-Supervised Training in Text-To-Speech
Python
169
star
9

STYLER

Official repository of STYLER: Style Factor Modeling with Rapidity and Robustness via Speech Decomposition for Expressive and Controllable Neural Text to Speech, INTERSPEECH 2021
Python
150
star
10

Comprehensive-E2E-TTS

A Non-Autoregressive End-to-End Text-to-Speech (text-to-wav), supporting a family of SOTA unsupervised duration modelings. This project grows with the research community, aiming to achieve the ultimate E2E-TTS
Python
140
star
11

Soft-DTW-Loss

PyTorch implementation of Soft-DTW: a Differentiable Loss Function for Time-Series in CUDA
Python
113
star
12

FastPitchFormant

PyTorch Implementation of NCSOFT's FastPitchFormant: Source-filter based Decomposed Modeling for Speech Synthesis
Python
70
star
13

VAENAR-TTS

PyTorch Implementation of VAENAR-TTS: Variational Auto-Encoder based Non-AutoRegressive Text-to-Speech Synthesis.
Python
69
star
14

WaveGrad2

PyTorch Implementation of Google Brain's WaveGrad 2: Iterative Refinement for Text-to-Speech Synthesis
Python
66
star
15

Daft-Exprt

PyTorch Implementation of Daft-Exprt: Robust Prosody Transfer Across Speakers for Expressive Speech Synthesis
Python
54
star
16

Comprehensive-Tacotron2

PyTorch Implementation of Google's Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions. This implementation supports both single-, multi-speaker TTS and several techniques to enforce the robustness and efficiency of the model.
Python
42
star
17

Robust_Fine_Grained_Prosody_Control

PyTorch Implementation of Robust and fine-grained prosody control of end-to-end speech synthesis
Python
39
star
18

Stepwise_Monotonic_Multihead_Attention

PyTorch Implementation of Stepwise Monotonic Multihead Attention similar to Enhancing Monotonicity for Robust Autoregressive Transformer TTS
Python
27
star
19

Deep-Learning-TTS-Template

This is a template for the Non-autoregressive Deep Learning-Based TTS model (in PyTorch).
Python
14
star
20

tacotron2_MMI

Another PyTorch implementation of Tacotron2 MMI (with waveglow) which supports n_frames_per_step>1 mode(reduction windows) and diagonal guided attention for robust alignments.
Jupyter Notebook
5
star
21

Fully_Hierarchical_Fine_Grained_TTS

Pytorch Implementation of Fully-hierarchical fine-grained prosody modeling for interpretable speech synthesis (Unofficial)
2
star
22

cs231n

cs231n 2020 Spring assignments implementation
Jupyter Notebook
2
star
23

pintos

KAIST CS330 OS pintos Project
HTML
1
star