These are the files used in the paper "A Machine Learning Approach to the Detection of Ghosting and Scattered Light Artifacts in Dark Energy Survey Images" (https://arxiv.org/abs/2105.10524)

Note:
  test and train sets consist of three numpy format files:
  x_*: contains the 400x400 low resolution images of the DECam focal plane
  y_*: contains the label (1:ghost or 0:not_ghost) for each image in "x"
  z_*: contains expnum, year, and filter for each image in "x"

./train_set: training set
  x_ghstcln400-cnn5_ins4.npy
  y_ghstcln400-cnn5_fix9z0_ins4.npy
  z_ghstcln400-cnn5_ins4.npy

./test_set: test set
  x_ghstcln400-tstdat.npy
  y_ghstcln400-tstdat.npy
  z_ghstcln400-tstdat.npy