Precise Recovery of Latent Vectors from Generative Adversarial Networks
This provides a practical solution for researchers and practitioners needing to invert GANs, though it appears incremental as it builds on existing gradient-based approaches.
The paper tackles the problem of reversing GAN mappings to recover latent vectors from generated images, achieving 100% recovery accuracy on GAN-generated images and demonstrating robustness to noise.
Generative adversarial networks (GANs) transform latent vectors into visually plausible images. It is generally thought that the original GAN formulation gives no out-of-the-box method to reverse the mapping, projecting images back into latent space. We introduce a simple, gradient-based technique called stochastic clipping. In experiments, for images generated by the GAN, we precisely recover their latent vector pre-images 100% of the time. Additional experiments demonstrate that this method is robust to noise. Finally, we show that even for unseen images, our method appears to recover unique encodings.