add a mode to generate pictures recursively

yenchenlin / pix2pix-tensorflow

TensorFlow implementation of "Image-to-Image Translation Using Conditional Adversarial Networks".

MIT License

939 stars 300 forks source link

add a mode to generate pictures recursively #36

Open dh7 opened 6 years ago

dh7 commented 6 years ago

Hi!

The following pull request update the code to add 3 feature: 1) the mode "one" allow the user to process any picture when the training is done 2) the mode "recursion" allow the user to use pix2pix recursively, to generate video based on next frame prediction 3) the mode "pre-recursion" allow the user to generate some image recursively from a set of frame