Generative Adversarial Networks with Inverse Transformation Unit
This is an incremental improvement for image generation and restoration tasks in computer vision.
The paper tackles the problem of improving Generative Adversarial Networks by introducing an inverse transformation unit, showing that it can sharpen and recover blurred images with certain transformations, as verified by sharpness measurements on MNIST and Fashion-MNIST datasets.
In this paper we introduce a new structure to Generative Adversarial Networks by adding an inverse transformation unit behind the generator. We present two theorems to claim the convergence of the model, and two conjectures to nonideal situations when the transformation is not bijection. A general survey on models with different transformations was done on the MNIST dataset and the Fashion-MNIST dataset, which shows the transformation does not necessarily need to be bijection. Also, with certain transformations that blurs an image, our model successfully learned to sharpen the images and recover blurred images, which was additionally verified by our measurement of sharpness.