“An in-depth assessment of convolutional neural networks for rail surface defect detection”, RSD, vol. 11, no. 8, p. e12211830252, Jun. 2022, doi: 10.33448/rsd-v11i8.30252.