JiahuiYu / generative_inpainting

DeepFill v1/v2 with Contextual Attention and Gated Convolution, CVPR 2018, and ICCV 2019 Oral
http://jiahuiyu.com/deepfill/
Other
3.27k stars 787 forks source link

In the flist, how is the image with missing parts and the true image passed? #555

Open GedorN opened 2 months ago

GedorN commented 2 months ago

Hello! First of all, thank you for the project! I am trying to train with my own dataset and I am facing a problem. I didn't understand how the training works in the flist structure. How does the training receive the image with the missing part and the true image in order to learn the difference? I understand that this file contains the path to the images, but how does it know which is which?