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

Last update: Jan 02, 2023

Overview

FastPitchFormant - PyTorch Implementation

PyTorch Implementation of FastPitchFormant: Source-filter based Decomposed Modeling for Speech Synthesis.

Quickstart

Dependencies

You can install the Python dependencies with

pip3 install -r requirements.txt

Inference

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

For English single-speaker TTS, run

python3 synthesize.py --text "YOUR_DESIRED_TEXT" --restore_step 1000000 --mode single -p config/LJSpeech/preprocess.yaml -m config/LJSpeech/model.yaml -t config/LJSpeech/train.yaml

The generated utterances will be put in output/result/.

Batch Inference

Batch inference is also supported, try

python3 synthesize.py --source preprocessed_data/LJSpeech/val.txt --restore_step 1000000 --mode batch -p config/LJSpeech/preprocess.yaml -m config/LJSpeech/model.yaml -t config/LJSpeech/train.yaml

to synthesize all utterances in preprocessed_data/LJSpeech/val.txt

Controllability

The pitch/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 pitch by 20 % by

python3 synthesize.py --text "YOUR_DESIRED_TEXT" --restore_step 1000000 --mode single -p config/LJSpeech/preprocess.yaml -m config/LJSpeech/model.yaml -t config/LJSpeech/train.yaml --duration_control 0.8 --pitch_control 0.8

Training

Datasets

The supported datasets are

LJSpeech: a single-speaker English dataset consists of 13100 short audio clips of a female speaker reading passages from 7 non-fiction books, approximately 24 hours in total.

Preprocessing

First, run

python3 prepare_align.py config/LJSpeech/preprocess.yaml

for some preparations.

As described in the paper, Montreal Forced Aligner (MFA) is used to obtain the alignments between the utterances and the phoneme sequences. Alignments for the LJSpeech datasets are provided here. You have to unzip the files in preprocessed_data/LJSpeech/TextGrid/.

After that, run the preprocessing script by

python3 preprocess.py config/LJSpeech/preprocess.yaml

Alternately, you can align the corpus by yourself. Download the official MFA package and run

./montreal-forced-aligner/bin/mfa_align raw_data/LJSpeech/ lexicon/librispeech-lexicon.txt english preprocessed_data/LJSpeech

./montreal-forced-aligner/bin/mfa_train_and_align raw_data/LJSpeech/ lexicon/librispeech-lexicon.txt preprocessed_data/LJSpeech

to align the corpus and then run the preprocessing script.

python3 preprocess.py config/LJSpeech/preprocess.yaml

Training

Train your model with

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

TensorBoard

Use

tensorboard --logdir output/log/LJSpeech

to serve TensorBoard on your localhost.

Implementation Issues

Use HiFi-GAN instead of VocGAN for vocoding.

Citation

@misc{lee2021fastpitchformant,
  author = {Lee, Keon},
  title = {FastPitchFormant},
  year = {2021},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/keonlee9420/FastPitchFormant}}
}

References

PyTorch implementation of Tacotron speech synthesis model.

tacotron_pytorch PyTorch implementation of Tacotron speech synthesis model. Inspired from keithito/tacotron. Currently not as much good speech quality

279 Dec 9, 2022

PyTorch implementation of Lip to Speech Synthesis with Visual Context Attentional GAN (NeurIPS2021)

Lip to Speech Synthesis with Visual Context Attentional GAN This repository contains the PyTorch implementation of the following paper: Lip to Speech

6 Nov 2, 2022

Implementation of Diverse Semantic Image Synthesis via Probability Distribution Modeling

Diverse Semantic Image Synthesis via Probability Distribution Modeling (CVPR 2021) Paper Zhentao Tan, Menglei Chai, Dongdong Chen, Jing Liao, Qi Chu,

45 Nov 17, 2022

A Flow-based Generative Network for Speech Synthesis

WaveGlow: a Flow-based Generative Network for Speech Synthesis Ryan Prenger, Rafael Valle, and Bryan Catanzaro In our recent paper, we propose WaveGlo

2k Dec 26, 2022

Official implementation of the paper: "LDNet: Unified Listener Dependent Modeling in MOS Prediction for Synthetic Speech"

LDNet Author: Wen-Chin Huang (Nagoya University) Email: [email protected] This is the official implementation of the paper "LDNet

40 Nov 20, 2022

TalkNet 2: Non-Autoregressive Depth-Wise Separable Convolutional Model for Speech Synthesis with Explicit Pitch and Duration Prediction.

TalkNet 2 [WIP] TalkNet 2: Non-Autoregressive Depth-Wise Separable Convolutional Model for Speech Synthesis with Explicit Pitch and Duration Predictio

69 Dec 17, 2022

HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis

HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis Jungil Kong, Jaehyeon Kim, Jaekyoung Bae In our paper, we p

31 Dec 8, 2022

BDDM: Bilateral Denoising Diffusion Models for Fast and High-Quality Speech Synthesis

Bilateral Denoising Diffusion Models (BDDMs) This is the official PyTorch implementation of the following paper: BDDM: BILATERAL DENOISING DIFFUSION M

172 Dec 23, 2022

This program uses trial auth token of Azure Cognitive Services to do speech synthesis for you.

🗣️ aspeak A simple text-to-speech client using azure TTS API(trial). 😆 TL;DR: This program uses trial auth token of Azure Cognitive Services to do s

359 Jan 5, 2023

Comments

Error when duration_control is <1

I can set any value above 1 (ie. '--duration_control 1.9') to slow down the speaking rate, but can't do the opposite, anything below 1 (ie. 0.9) will throw this error message:

C:\FastPitchFormant>python synthesize.py --text "testing" --restore_step 600000 --mode single -p config/LJSpeech/preprocess.yaml -m config/LJSpeech/model.yaml -t config/LJSpeech/train.yaml --duration_control 0.9 --pitch_control 1
Removing weight norm...
Raw Text Sequence: testing
Phoneme Sequence: {T EH1 S T IH0 NG}
Traceback (most recent call last):
  File "synthesize.py", line 207, in <module>
    synthesize(model, args.restore_step, configs, vocoder, batchs, control_values)
  File "synthesize.py", line 95, in synthesize
    output = model(
  File "C:\ProgramData\Anaconda3\envs\pttf2cu111py38\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "C:\FastPitchFormant\model\FastPitchFormant.py", line 89, in forward
    formant_hidden = self.formant_generator(h, mel_masks)
  File "C:\ProgramData\Anaconda3\envs\pttf2cu111py38\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "C:\FastPitchFormant\model\modules.py", line 329, in forward
    output, enc_slf_attn = enc_layer(
  File "C:\ProgramData\Anaconda3\envs\pttf2cu111py38\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "C:\FastPitchFormant\model\blocks.py", line 109, in forward
    enc_output, enc_slf_attn = self.slf_attn(
  File "C:\ProgramData\Anaconda3\envs\pttf2cu111py38\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "C:\FastPitchFormant\model\blocks.py", line 162, in forward
    output, attn = self.attention(q, k, v, mask=mask)
  File "C:\ProgramData\Anaconda3\envs\pttf2cu111py38\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "C:\FastPitchFormant\model\blocks.py", line 189, in forward
    attn = attn.masked_fill(mask, -np.inf)
RuntimeError: The size of tensor a (32) must match the size of tensor b (34) at non-singleton dimension 2

opened by MaxGodTier 1

Releases(v1.0.0)

v1.0.0(Aug 3, 2021)

First Official Release
Source code(tar.gz)
Source code(zip)
v0.1.0(Jul 17, 2021)

Source code(tar.gz)
Source code(zip)

Owner

Keon Lee

GitHub Repository

中文语音识别系列，读者可以借助它快速训练属于自己的中文语音识别模型，或直接使用预训练模型测试效果。

MASR中文语音识别(pytorch版) 开箱即用自行训练使用与训练分离(增量训练) 识别率高说明：因为每个人电脑机器不同，而且有些安装包安装起来比较麻烦，强烈建议直接用我编译好的docker环境跑目前docker基础环境为ubuntu-cuda10.1-cudnn7-pytorch1.6.

180 Dec 17, 2022

A simple baseline for the 2022 IEEE GRSS Data Fusion Contest (DFC2022)

DFC2022 Baseline A simple baseline for the 2022 IEEE GRSS Data Fusion Contest (DFC2022) This repository uses TorchGeo, PyTorch Lightning, and Segmenta

24 Nov 28, 2022

SMIS - Semantically Multi-modal Image Synthesis(CVPR 2020)

Semantically Multi-modal Image Synthesis Project page / Paper / Demo Semantically Multi-modal Image Synthesis(CVPR2020). Zhen Zhu, Zhiliang Xu, Anshen

316 Dec 01, 2022

Automatically erase objects in the video, such as logo, text, etc.

Video-Auto-Wipe Read English Introduction：Here 本人不定期的基于生成技术制作一些好玩有趣的算法模型，这次带来的作品是“视频擦除”方向的应用模型，它实现的功能是自动感知到视频中我们不想看见的部分（譬如广告、水印、字幕、图标等等）然后进行擦除。由于图标擦

141 Dec 26, 2022

Listing arxiv - Personalized list of today's articles from ArXiv

Personalized list of today's articles from ArXiv Print and/or send to your gmail

5 Jun 17, 2022

Baseline powergrid model for NY

Baseline-powergrid-model-for-NY Table of Contents About The Project Built With Usage License Contact Acknowledgements About The Project As the urgency

6 Nov 24, 2022

Fantasy Points Prediction and Dream Team Formation

Fantasy-Points-Prediction-and-Dream-Team-Formation Collected Data from open source resources that have over 100 Parameters for predicting cricket play

2 Sep 13, 2022

UNAVOIDS: Unsupervised and Nonparametric Approach for Visualizing Outliers and Invariant Detection Scoring

UNAVOIDS: Unsupervised and Nonparametric Approach for Visualizing Outliers and Invariant Detection Scoring Code Summary aggregate.py: this script aggr

1 Dec 28, 2021

Ejemplo Algoritmo Viterbi - Example of a Viterbi algorithm applied to a hidden Markov model on DNA sequence

Ejemplo Algoritmo Viterbi Ejemplo de un algoritmo Viterbi aplicado a modelo ocul

1 Jan 10, 2022

This is the dataset and code release of the OpenRooms Dataset.

95 Jan 08, 2023

TCube generates rich and fluent narratives that describes the characteristics, trends, and anomalies of any time-series data (domain-agnostic) using the transfer learning capabilities of PLMs.

TCube: Domain-Agnostic Neural Time series Narration This repository contains the code for the paper: "TCube: Domain-Agnostic Neural Time series Narrat

7 Oct 31, 2021

[AAAI2022] Source code for our paper《Suppressing Static Visual Cues via Normalizing Flows for Self-Supervised Video Representation Learning》

SSVC The source code for paper [Suppressing Static Visual Cues via Normalizing Flows for Self-Supervised Video Representation Learning] samples of the

7 Oct 26, 2022

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

Related tags

Overview

FastPitchFormant - PyTorch Implementation

Quickstart

Dependencies

Inference

Batch Inference

Controllability

Training

Datasets

Preprocessing

Training

TensorBoard

Implementation Issues

Citation

References

You might also like...

PyTorch implementation of Tacotron speech synthesis model.

PyTorch implementation of Lip to Speech Synthesis with Visual Context Attentional GAN (NeurIPS2021)

Implementation of Diverse Semantic Image Synthesis via Probability Distribution Modeling

A Flow-based Generative Network for Speech Synthesis

Official implementation of the paper: "LDNet: Unified Listener Dependent Modeling in MOS Prediction for Synthetic Speech"

TalkNet 2: Non-Autoregressive Depth-Wise Separable Convolutional Model for Speech Synthesis with Explicit Pitch and Duration Prediction.

HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis

BDDM: Bilateral Denoising Diffusion Models for Fast and High-Quality Speech Synthesis

This program uses trial auth token of Azure Cognitive Services to do speech synthesis for you.

Comments

Error when duration_control is <1

Releases(v1.0.0)

v1.0.0(Aug 3, 2021)

v0.1.0(Jul 17, 2021)

Owner

Keon Lee

中文语音识别系列，读者可以借助它快速训练属于自己的中文语音识别模型，或直接使用预训练模型测试效果。

A simple baseline for the 2022 IEEE GRSS Data Fusion Contest (DFC2022)

SMIS - Semantically Multi-modal Image Synthesis(CVPR 2020)

Automatically erase objects in the video, such as logo, text, etc.

Listing arxiv - Personalized list of today's articles from ArXiv

Baseline powergrid model for NY

Fantasy Points Prediction and Dream Team Formation

UNAVOIDS: Unsupervised and Nonparametric Approach for Visualizing Outliers and Invariant Detection Scoring

Ejemplo Algoritmo Viterbi - Example of a Viterbi algorithm applied to a hidden Markov model on DNA sequence

This is the dataset and code release of the OpenRooms Dataset.

TCube generates rich and fluent narratives that describes the characteristics, trends, and anomalies of any time-series data (domain-agnostic) using the transfer learning capabilities of PLMs.

[AAAI2022] Source code for our paper《Suppressing Static Visual Cues via Normalizing Flows for Self-Supervised Video Representation Learning》

Coded illumination for improved lensless imaging

Implementation of the Chamfer Distance as a module for pyTorch

A PyTorch library for Vision Transformers

Keras-tensorflow implementation of Fully Convolutional Networks for Semantic Segmentation（Unfinished）

Codebase for "ProtoAttend: Attention-Based Prototypical Learning."

Implementation of CVPR 2021 paper "Spatially-invariant Style-codes Controlled Makeup Transfer"

Semi-supervised Representation Learning for Remote Sensing Image Classification Based on Generative Adversarial Networks

🤗 Paper Style Guide