wav2vec 2.0 with CTC/Attention trained on Switchboard (No LM)

This repository provides all the necessary tools to perform automatic speech recognition from an end-to-end system pretrained on the Switchboard (EN) corpus within SpeechBrain. For a better experience, we encourage you to learn more about SpeechBrain.

The performance of the model is the following:

Release	Swbd CER	Callhome CER	Eval2000 CER	Swbd WER	Callhome WER	Eval2000 WER	GPUs
17-09-22	5.24	9.69	7.44	8 .76	14.67	11.78	4xA100 40GB

Pipeline Description

This ASR system is composed of 2 different but linked blocks:

Tokenizer (unigram) that transforms words into subword units trained on the Switchboard training transcriptions and the Fisher corpus.
Acoustic model (wav2vec2.0 + CTC). A pretrained wav2vec 2.0 model (facebook/wav2vec2-large-lv60) is combined with a feature encoder consisting of three DNN layers and finetuned on Switchboard. The obtained final acoustic representation is given to a greedy CTC decoder.

The system is trained with recordings sampled at 16kHz (single channel).
The code will automatically normalize your audio (i.e., resampling + mono channel selection) when calling transcribe_file if needed.

Install SpeechBrain

First of all, please install tranformers and SpeechBrain with the following command:

pip install speechbrain transformers

Please notice that we encourage you to read our tutorials and learn more about SpeechBrain.

Transcribing Your Own Audio Files

from speechbrain.inference.ASR import EncoderASR

asr_model = EncoderASR.from_hparams(source="speechbrain/asr-wav2vec2-switchboard", savedir="pretrained_models/asr-wav2vec2-switchboard")
asr_model.transcribe_file('speechbrain/asr-wav2vec2-switchboard/example.wav')

Inference on GPU

To perform inference on the GPU, add run_opts={"device":"cuda"} when calling the from_hparams method.

Training

The model was trained with SpeechBrain (commit hash: 70904d0). To train it from scratch follow these steps:

Clone SpeechBrain:

git clone https://github.com/speechbrain/speechbrain/

Install it:

cd speechbrain
pip install -r requirements.txt
pip install -e .

Run Training:

cd recipes/Switchboard/ASR/CTC
python train_with_wav2vec.py hparams/train_with_wav2vec.yaml --data_folder=your_data_folder

Limitations

The SpeechBrain team does not provide any warranty on the performance achieved by this model when used on other datasets.

Credits

This model was trained with resources provided by the THN Center for AI.

About SpeechBrain

SpeechBrain is an open-source and all-in-one speech toolkit. It is designed to be simple, extremely flexible, and user-friendly. Competitive or state-of-the-art performance is obtained in various domains.

Website: https://speechbrain.github.io/
GitHub: https://github.com/speechbrain/speechbrain/
HuggingFace: https://huggingface.co/speechbrain/

Citing SpeechBrain

Please cite SpeechBrain if you use it for your research or business.

@misc{speechbrain,
  title={{SpeechBrain}: A General-Purpose Speech Toolkit},
  author={Mirco Ravanelli and Titouan Parcollet and Peter Plantinga and Aku Rouhe and Samuele Cornell and Loren Lugosch and Cem Subakan and Nauman Dawalatabad and Abdelwahab Heba and Jianyuan Zhong and Ju-Chieh Chou and Sung-Lin Yeh and Szu-Wei Fu and Chien-Feng Liao and Elena Rastorgueva and François Grondin and William Aris and Hwidong Na and Yan Gao and Renato De Mori and Yoshua Bengio},
  year={2021},
  eprint={2106.04624},
  archivePrefix={arXiv},
  primaryClass={eess.AS},
  note={arXiv:2106.04624}
}