https://github.com/haoxiangsnr/A-Convolutional-Recurrent-Neural-Network-for-Real-Time-Speech-Enhancement

A minimum unofficial implementation of the "A Convolutional Recurrent Neural Network for Real-Time Speech Enhancement" (CRN) using PyTorch
https://github.com/haoxiangsnr/A-Convolutional-Recurrent-Neural-Network-for-Real-Time-Speech-Enhancement

cnn cnn-rnn pytorch real-time rnn speech-enhancement speech-processing

Last synced: about 2 months ago
JSON representation

A minimum unofficial implementation of the "A Convolutional Recurrent Neural Network for Real-Time Speech Enhancement" (CRN) using PyTorch

• Host: GitHub
• URL: https://github.com/haoxiangsnr/A-Convolutional-Recurrent-Neural-Network-for-Real-Time-Speech-Enhancement
• Owner: haoxiangsnr
• Created: 2019-08-12T08:02:05.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2020-09-05T00:50:59.000Z (about 4 years ago)
• Last Synced: 2024-05-16T12:58:25.953Z (4 months ago)
• Topics: cnn, cnn-rnn, pytorch, real-time, rnn, speech-enhancement, speech-processing
• Language: Python
• Homepage:
• Size: 43.9 KB
• Stars: 293
• Watchers: 7
• Forks: 53
• Open Issues: 9
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# A Convolutional Recurrent Neural Network for Real-Time Speech Enhancement

A minimum unofficial implementation of the [A Convolutional Recurrent Neural Network for Real-Time Speech Enhancement (CRN)](https://www.isca-speech.org/archive/Interspeech_2018/pdfs/1405.pdf) using PyTorch.

## ToDo

- [x] Real-time version

- [x] Update trainer

- [x] Visualization of the spectrogram and the metrics (PESQ, STOI, SI-SDR) in the training

- [ ] More docs

## Usage

Training:

```

python train.py -C config/train/baseline_model.json5

```

Inference:

```

python inference.py \

    -C config/inference/basic.json5 \

    -cp ~/Experiments/CRN/baseline_model/checkpoints/latest_model.tar \

    -dist ./enhanced

```

Check out the README of [Wave-U-Net for SE](https://github.com/haoxiangsnr/Wave-U-Net-for-Speech-Enhancement) to learn more.

## Performance

PESQ, STOI, SI-SDR on DEMAND - Voice Bank test dataset, for reference only:

| Experiment | PESQ | SI-SDR | STOI |

| --- | --- | --- | --- |

|Noisy | 1.979 | 8.511| 0.9258|

|CRN | 2.528| 17.71| 0.9325|

|CRN signal approximation  |2.606 |17.84 |0.9382|

## Dependencies

- Python==3.\*.\*

- torch==1.\*

- librosa==0.7.0

- tensorboard

- pesq

- pystoi

- matplotlib

- tqdm

## References

- [CRNN_mapping_baseline](https://github.com/YangYang/CRNN_mapping_baseline)

- [A Convolutional Recurrent Neural Network for Real-Time Speech Enhancement](https://web.cse.ohio-state.edu/~wang.77/papers/Tan-Wang1.interspeech18.pdf)

- [EHNet](https://github.com/ododoyo/EHNet)

- [Convolutional-Recurrent Neural Networks for Speech Enhancement](https://arxiv.org/abs/1805.00579)