This is a Audio enhancement system which uses GAN.
Data: https://datashare.is.ed.ac.uk/bitstream/handle/10283/2791/
How to run the code? py data_preprocess.py for data pre-processing
py main.py for model training
after every epoch .pkl checkpoints of both generator and discriminator is saved to the local directory.
py test_audio.py for enhancing the audio you can record the audio up to 1 minutes(can be changed to desired time span from record.py file) OR you can feed the recorded audio into the system and also enhance.(both can be changed inside the test_audio.py file)
also can select the generator checkpoint i.e generator-100.pkl is set to default here but again you can change it in (you know where..)
ENJOY!!!!!!!