crisbodnar / text-to-image

Text to Image Synthesis using Generative Adversarial Networks
https://arxiv.org/abs/1805.00676
76 stars 28 forks source link

code to process text pickle and data split #8

Open khcy82dyc opened 5 years ago

khcy82dyc commented 5 years ago

hi, thanks for the great work. just wondering if there is a code available to generate char-CNN-RNN-embeddings.pickle class_info.pickle and filenames.pickle and split it into train and test? I'm hoping to test it on my own datasets.. thanks!

crisbodnar commented 5 years ago

Sorry for the late reply, @khcy82dyc ! You can find some instructions on this github page: https://github.com/reedscot/cvpr2016 . Unfortunately, the scripts are written in lua and they use torch, but you could check the web for implementations in other frameworks or just do it yourself.

saba56 commented 2 years ago

Have you @khcy82dyc been able to generate class_info.pickle and filenames.pickle.