I am new to Deep Learning and Generative Adversarial Network. I want to train pix2pixHD GAN on my own images but I can not understand the procedure for creating my own dataset.