Pretrained model for Colorization
HaHaFish1014 opened this issue · 2 comments
Thank you for your great work and share your code!
Would you like to share the pretained model of colorization?
Hello, I would like to use this model to colorize images. May I ask which images should be placed in each of the three folders in the dataset? After the training is completed, what are the purposes of the files that appear in the result folder? How can I obtain the colorized images?
Hello, I would like to use this model to colorize images. May I ask which images should be placed in each of the three folders in the dataset? After the training is completed, what are the purposes of the files that appear in the result folder? How can I obtain the colorized images?
Sorry for not responding in time.
About the three folders in the dataset, the author has elaborated in the Data preparation. I can give you a sample.
/BBDM-main/data/train/A/trainGray.png
/BBDM-main/data/train/B/trainColor.png
/BBDM-main/data/test/A/testGray.png
/BBDM-main/data/test/B/testColor.png
/BBDM-main/data/val/A/valGray.png
/BBDM-main/data/val/B/valColor.png
After training, you will get your own trained_model in the results folder. The checkpoint will save in results/XXXX/LBBDM-f8/checkpoint. You can use the checkpoint to sample some images, which will save in results/XXXX/LBBDM-f8/sample_to_eval.