This repository implements Zachary C. Lipton and Subarna Tripathi's paper Precise Recovery of Latent Vectors from Generative Adversarial Networks, ICLR 2017 workshop track [arxiv]. https://arxiv.org/abs/1702.04782
Most of the code in this repository was written by modifying Brandon Amos's DCGAN-completion-tensorflow project, which is MIT-licensed. Our modifications are also MIT-licensed. The ./checkpoint_new2 directory contains a pre-trained model for faces used in the paper.
Yixing Lao's implementation in Pytorch: https://github.com/yxlao/pytorch-reverse-gan