DenoiseGAN A GAN-Based Method for seismic signal denoising

1.intall

1.1 if you have already installed Git LFS

git lfs clone --depth=1 https://github.com/zpv2jdfc/DenoiseGAN.git DenoiseGAN

1.2 otherwise

git clone --depth=1 -b withoutmodel https://github.com/zpv2jdfc/DenoiseGAN.git DenoiseGAN
download model and put it into DenoiseGAN/gan/models

(If you haven't used git and download the zip file directly, you still need to download model and put it into DenoiseGAN/gan/models)

. open directory

cd DenoiseGAN

. install & activate DenoiseGAN environment
conda env create -f env.yml

conda activate denoisegan

. Drop your files or directories to gan/test_data. (only support .mseed or .h5 file now)

. Run command python main.py. The denoised signal files will be placed in gan/results

. If you also want to see the figures of spectrum, use python main.py true. But this will be slow.