End-to-End Conditional GAN-based Architectures for Image Colourisation
This work addresses the issue of lack of colorfulness in image colorization for computer vision applications, but it is incremental as it builds on existing GAN and U-Net frameworks.
The paper tackled the problem of generating realistic colors for grayscale images by developing an end-to-end conditional GAN architecture, achieving improved colorization results on the ILSVRC 2012 dataset compared to other GAN-based methods.
In this work recent advances in conditional adversarial networks are investigated to develop an end-to-end architecture based on Convolutional Neural Networks (CNNs) to directly map realistic colours to an input greyscale image. Observing that existing colourisation methods sometimes exhibit a lack of colourfulness, this paper proposes a method to improve colourisation results. In particular, the method uses Generative Adversarial Neural Networks (GANs) and focuses on improvement of training stability to enable better generalisation in large multi-class image datasets. Additionally, the integration of instance and batch normalisation layers in both generator and discriminator is introduced to the popular U-Net architecture, boosting the network capabilities to generalise the style changes of the content. The method has been tested using the ILSVRC 2012 dataset, achieving improved automatic colourisation results compared to other methods based on GANs.