Giters

YoshikiMas / s3prl

Audio Foundation Models (Self-Supervised Speech/Sound Pre-training and Representation Learning Toolkit)

Home Page:https://s3prl.github.io/s3prl/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool



Apache License 2.0 CC_BY_NC License CI Codecov Bitbucket open issues

Notice for pull requests

Please first discuss with us on the issue page about your feature request before implementing the actual pull request, so we can discuss about how to achieve the functionality. If we did not discuss about the detail, it is highly possible that we are not accepting the pull request due to the difficulty of maintenance.

Environment compatibilities CI

We support the following environments. The test cases are ran with tox locally and on github action:

Env versions
os ubuntu-18.04, ubuntu-20.04
python 3.7, 3.8, 3.9, 3.10
pytorch 1.8.1, 1.9.1, 1.10.2, 1.11.0, 1.12.1

Change Log

Please note that

As the core developers and repository owners, Andy T. Liu and Shu-wen (Leo) Yang never forced others to contribute to the S3PRL project. Instead, most of the times people want and ask to contribute to this proejct. Feel free to ask old contributors to check. We built this project from scratch and developed most of the functions ourselves since 2019 and received huge amount of stars. There were around 400 stars before we started to develop SUPERB Benchmark. (Not to mention that we also recruited most of the first generation SUPERB team members at National Taiwan University ourselves and led the development of SUPERB and rebuilt the entire downstream evaluation codebase. No one ever paid us for building this project which we started ourselves in the hope to help others and we did. There are several papers using our upstream collection function citing the S3PRL github link before SUPERB's existence. For example, the S2VC paper in the Used by section below.) Hence people were interested to contribute to our project and develop new functions together. From our perspective, we were kind enough to help reviewing PRs and help others build their resume. (We could just refuse them in the beginning.) But later, someone started to question about our ownership on this repository and manipulate the information, just because we allowed others to contribute. This made us really disappointed. We hence thought about never allowing new contributions in the future. If any old contributors felt unhappy about the current situation, it is okay to dicuss about removing your previous contributions, and we are happy to help. Our principle is always to give the biggest credit to those who already contributed (especially who fought with us when no one knew S3PRL,) instead of someone wishing to take over the repository by undiscussed and overwhelming PRs after the repository is successful. We still welcome new contributions as long as the development is fully discussed and the new contributors accept our principle above. We are grateful and care about every precious old contributions and refuse to dilute old contributors' credit by accepting lots of new PRs. Please find the names below to know the people who put huge effort on making this project successful.

As a worth-mentioning point, we are not accepting new SUPERB development in the future. There are two major reasons:

  • The SUPERB team members changed a lot and rapidly. Andy and Leo are not available to involve every new effort, hence we are afraid to accept undiscussed and unreviewed while huge amount of new implementations. We won't get benefit from helping reviewing these efforts, while people will further question about our ownership and try to take over it, and the new members might not always want to contribute to our project. Hence, there is no benefit to both sides.
  • The current S3PRL is getting too large and becoming hard to maintain and keep the dependency small.

Hence, we will do our best in maintaining SUPERB and SUPERB-SG, and we believe it is better to let the new SUPERB members to maintain their own implementation in their own repositories, just like how we started from nothing when developing the first version of SUPERB. So both sides get proper credit to their own development.

Introduction and Usages

This is an open source toolkit called s3prl, which stands for Self-Supervised Speech Pre-training and Representation Learning. Self-supervised speech pre-trained models are called upstream in this toolkit, and are utilized in various downstream tasks.

The toolkit has three major usages:

Pretrain

  • Pretrain upstream models, including Mockingjay, Audio ALBERT and TERA.
  • Document: pretrain/README.md

Upstream

  • Easily load most of the existing upstream models with pretrained weights in a unified I/O interface.
  • Pretrained models are registered through torch.hub, which means you can use these models in your own project by one-line plug-and-play without depending on this toolkit's coding style.
  • Document: upstream/README.md

Downstream

Below is an intuitive illustration on how this toolkit may help you:



Feel free to use or modify our toolkit in your research. Here is a list of papers using our toolkit. Any question, bug report or improvement suggestion is welcome through opening up a new issue.

If you find this toolkit helpful to your research, please do consider citing our papers, thanks!

Installation

  1. Python >= 3.6
  2. Install sox on your OS
  3. Install s3prl: Read doc or pip install -e ".[all]"
  4. (Optional) Some upstream models require special dependencies. If you encounter error with a specific upstream model, you can look into the README.md under each upstream folder. E.g., upstream/pase/README.md

Development pattern for contributors

  1. Create a personal fork of the main S3PRL repository in GitHub.
  2. Make your changes in a named branch different from master, e.g. you create a branch new-awesome-feature.
  3. Contact us if you have any questions during development.
  4. Generate a pull request through the Web interface of GitHub.
  5. Please verify that your code is free of basic mistakes, we appreciate any contribution!

Reference Repositories

License

The majority of S3PRL Toolkit is licensed under the Apache License version 2.0, however all the files authored by Facebook, Inc. (which have explicit copyright statement on the top) are licensed under CC-BY-NC.

Used by

List of papers that used our toolkit (Feel free to add your own paper by making a pull request)

Self-Supervised Pretraining

Explanability

Adversarial Attack

Voice Conversion

Benchmark and Evaluation

  • SUPERB: Speech processing Universal PERformance Benchmark (Yang et al., 2021)

    @misc{superb,
      title={SUPERB: Speech processing Universal PERformance Benchmark}, 
      author={Shu-wen Yang and Po-Han Chi and Yung-Sung Chuang and Cheng-I Jeff Lai and Kushal Lakhotia and Yist Y. Lin and Andy T. Liu and Jiatong Shi and Xuankai Chang and Guan-Ting Lin and Tzu-Hsien Huang and Wei-Cheng Tseng and Ko-tik Lee and Da-Rong Liu and Zili Huang and Shuyan Dong and Shang-Wen Li and Shinji Watanabe and Abdelrahman Mohamed and Hung-yi Lee},
      year={2021},
      eprint={2105.01051},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

  • Utilizing Self-supervised Representations for MOS Prediction (Tseng et al., 2021)

    @misc{ssr_mos,
    title={Utilizing Self-supervised Representations for MOS Prediction}, 
    author={Wei-Cheng Tseng and Chien-yu Huang and Wei-Tsung Kao and Yist Y. Lin and Hung-yi Lee},
    year={2021},
    eprint={2104.03017},
    archivePrefix={arXiv},
    primaryClass={eess.AS}
}

}

Citation

If you find this toolkit useful, please consider citing following papers.

  • If you use our pre-training scripts, or the downstream tasks considered in TERA and Mockingjay, please consider citing the following:
@misc{tera,
  title={TERA: Self-Supervised Learning of Transformer Encoder Representation for Speech},
  author={Andy T. Liu and Shang-Wen Li and Hung-yi Lee},
  year={2020},
  eprint={2007.06028},
  archivePrefix={arXiv},
  primaryClass={eess.AS}
}

@article{mockingjay,
   title={Mockingjay: Unsupervised Speech Representation Learning with Deep Bidirectional Transformer Encoders},
   ISBN={9781509066315},
   url={http://dx.doi.org/10.1109/ICASSP40776.2020.9054458},
   DOI={10.1109/icassp40776.2020.9054458},
   journal={ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
   publisher={IEEE},
   author={Liu, Andy T. and Yang, Shu-wen and Chi, Po-Han and Hsu, Po-chun and Lee, Hung-yi},
   year={2020},
   month={May}
}
  • If you use our organized upstream interface and features, or the SUPERB downstream benchmark, please consider citing the following:
@inproceedings{yang21c_interspeech,
  author={Shu-wen Yang and Po-Han Chi and Yung-Sung Chuang and Cheng-I Jeff Lai and Kushal Lakhotia and Yist Y. Lin and Andy T. Liu and Jiatong Shi and Xuankai Chang and Guan-Ting Lin and Tzu-Hsien Huang and Wei-Cheng Tseng and Ko-tik Lee and Da-Rong Liu and Zili Huang and Shuyan Dong and Shang-Wen Li and Shinji Watanabe and Abdelrahman Mohamed and Hung-yi Lee},
  title={{SUPERB: Speech Processing Universal PERformance Benchmark}},
  year=2021,
  booktitle={Proc. Interspeech 2021},
  pages={1194--1198},
  doi={10.21437/Interspeech.2021-1775}
}

About

Audio Foundation Models (Self-Supervised Speech/Sound Pre-training and Representation Learning Toolkit)

https://s3prl.github.io/s3prl/

License:Apache License 2.0

Languages

Language:Python 98.0%Language:Shell 2.0%Language:Dockerfile 0.1%