GANspection
This work provides insights into GAN behavior for researchers in unsupervised learning, though it is incremental as it builds on existing inspection methods.
The paper tackles the problem of understanding the latent space representations learned by Generative Adversarial Networks (GANs) by using interpolation, extrapolation, and vector arithmetic techniques to inspect the learned manifold, showing that GANs learn a data probability distribution rather than memorizing images and encode semantically relevant information.
Generative Adversarial Networks (GANs) have been used extensively and quite successfully for unsupervised learning. As GANs don't approximate an explicit probability distribution, it's an interesting study to inspect the latent space representations learned by GANs. The current work seeks to push the boundaries of such inspection methods to further understand in more detail the manifold being learned by GANs. Various interpolation and extrapolation techniques along with vector arithmetic is used to understand the learned manifold. We show through experiments that GANs indeed learn a data probability distribution rather than memorize images/data. Further, we prove that GANs encode semantically relevant information in the learned probability distribution. The experiments have been performed on two publicly available datasets - Large Scale Scene Understanding (LSUN) and CelebA.