I want to train HyperGAN with a set of 400+ images of people, but they aren't the specified size (32x32 pixels) for training.


Is there any way/program to help cropping/resizing them so as to not do it 100% manually?


Cropping and/or resizing is very trivial using OpenCV. You could write a 10 line script to iterate over all your images and apply the transformation you require.

  • $\begingroup$ Just an advice, you can take it if you want to or choose to ignore it, but you won't learn much if you simply run codes published online, especially in such difficult topics like GANs and DL. Starting from the basics in Machine Learning/Deep Learning and CV through online courses, and then moving to such topics would be much more helpful. Just an opinion. $\endgroup$
    – tomar__
    Oct 29 '17 at 18:47
  • $\begingroup$ Thanks for the info, I've been trying to get into ML for a while, and this is the first time I got CUDA working so I wanted to do something, but I will develop my own stuff after this =) (Probably in C++ though) $\endgroup$
    – Teivel
    Oct 29 '17 at 20:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.