Text-to-Image-Synthesis
Text-to-Image-Synthesis copied to clipboard
New dataset
Can someone provide guidance on how to generate hdf5 file for COCO dataset. How do I get the following? birds_images_path: '/export/mlrg/aelnouby/projects/GANs/Birds dataset/CUB_200_2011/CUB_200_2011/images/' birds_embedding_path: '/export/mlrg/aelnouby/projects/GANs/Birds dataset/cub_icml/' birds_text_path: '/export/mlrg/aelnouby/projects/GANs/Birds dataset/cvpr2016_cub/text_c10/'
val_split_path: '/export/mlrg/aelnouby/projects/GANs/Birds dataset/cub_icml/valclasses.txt' train_split_path: '/export/mlrg/aelnouby/projects/GANs/Birds dataset/cub_icml/trainclasses.txt' test_split_path: '/export/mlrg/aelnouby/projects/GANs/Birds dataset/cub_icml/testclasses.txt'
There was some reference to use https://github.com/reedscot/cvpr2016, but when I checked the code, it didnt provide much insight. Any help would be appreciated
Check this repo https://github.com/reedscot/icml2016
I am not actually sure if everything you need is there, but it seems the caption embeddings and the image data links are there as well.