/DDAE_PyTorch

(Reproduced with PyTorch) : DDAE speech enhancement on spectrogram domain using Keras

Primary LanguagePython

DDAE speech enhancement

Hyper-parameters are not tuned to optimal

Fork

DDAE is reproduced with PyTorch in this repo.

Run with python3.9 and pytorch1.5

  • pip install -r v2_py39_torch19/requirements.txt
  • python v2_py39_torch19/v2_spectrum.py data.h5 list_noisy list_clean
  • python v2_py39_torch19/v2_train_DNN.py data.h5
  • python v2_py39_torch19/v2_test_gen_spec.py DDAE.pt list_noisy

Run with python2 and tensorflow 1

  • conda create --name py27 python=2.7
  • conda activate py27
  • pip install tensorflow=1.2
  • pip install keras=1.2.2
  • pip install librosa==0.3.1
  • pip install scikit-learn==0.16.1
  • then run commands below ("Getting Started")

Prerequisites:

  • Keras 1.2
  • Tensorflow 1.x as backend
  • h5py
  • librosa
  • scipy

Getting Started:

Extract spectrogram features:

python spectrum.py data.h5 list_noisy list_clean

Train DDAE using Keras:

python train_DNN.py data.h5

Enhance test wave files using trained model:

python test_gen_spec.py model.hdf5 list_noisy