Sorry, you need to enable JavaScript to visit this website.

End-to-End Conditional GAN-based Architectures for Image Colourisation

Citation Author(s):
Marc Gorriz Blanch, Marta Mrak, Alan Smeaton, Noel O'Connor
Submitted by:
Marc Gorriz Blanch
Last updated:
24 September 2019 - 12:11pm
Document Type:
Poster
Document Year:
2019
Event:
Presenters Name:
Marc Gorriz Blanch
Paper Code:
147

Abstract 

Abstract: 

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.

up
0 users have voted:

Dataset Files

MMSP2019_poster.pdf

(126)