iggcaswy / ConvDeNoise

A convolutional denoising autoencoder to denoise correlation functions

ConvDeNoise: A convolutional denoising autoencoder to denoise correlation functions

• Codes to reproduce Figure 7 (Figure 8 can also be plotted by changing two lines of the code) of the following paper:

  • Viens L. and Van Houtte C., Denoising ambient seismic field correlation functions with convolutional autoencoders (submitted to GJI). A preprint (not peer-reviewed) of the paper is available at https://eartharxiv.org/q4m2t/.

• The autoencoder is composed of an encoder part and a decoder part:

• The Codes folder contains 5 files:

  • The Reproduce_Fig_7.py file is the python code to reproduce Figure 7.
  • The functions_for_autoencoders.py file contains functions to bandpass filter the data with a Butterworth filter, denoise the SC functions with the SVDWF method (Moreau et al., 2017), and compute the stretching to retrieve dv/v measurements
  • The ConvDeNoise_NS7M_station.h5 contains the weights of ConvDeNoise trained for the NS7M station (Requires Keras 2.2.4)
  • The Test_data.mat contains 16 days of raw SC functions at the NS7M station, reference waveforms to compute the dv/v,... (e.g., all the data required to reproduce Figure 7).
  • The ConvDeNoise_core.py file is the convolutional denoising autoencoder main code that was used to compute the ConvDeNoise_NS7M_station.h5 file (requires the raw SC functions, please email me for the training set, file is too big for Github)