cGANs for Cartoon to Real-life Images

TL;DR

This paper explores the use of cGANs, specifically Pix2Pix, for translating cartoon images into realistic photographs, evaluating its robustness on cartoon-to-real datasets.

Contribution

It applies the Pix2Pix model to cartoon-to-real image translation and assesses its robustness in this specific domain.

Findings

Pix2Pix can generate realistic images from cartoon inputs.

The model's robustness varies across different datasets.

Conditional GANs are effective for image-to-image translation tasks.

Abstract

The image-to-image translation is a learning task to establish a visual mapping between an input and output image. The task has several variations differentiated based on the purpose of the translation, such as synthetic to real translation, photo to caricature translation, and many others. The problem has been tackled using different approaches, either through traditional computer vision methods, as well as deep learning approaches in recent trends. One approach currently deemed popular and effective is using the conditional generative adversarial network, also known shortly as cGAN. It is adapted to perform image-to-image translation tasks with typically two networks: a generator and a discriminator. This project aims to evaluate the robustness of the Pix2Pix model by applying the Pix2Pix model to datasets consisting of cartoonized images. Using the Pix2Pix model, it should be possible to train the network to generate real-life images from the cartoonized images.