# Resolution enhancement in the recovery of underdrawings via style transfer by generative adversarial deep neural networks

arXiv: https://arxiv.org/abs/2102.00209 · HTML: https://mlsanity.com/paper/2102.00209/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 19 references that are in the map.
- First posted on arXiv 30 Jan 2021.

## Builds on (most cited first, 8 of 8)

- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs](https://mlsanity.com/paper/1711.11585/) (2017)
- [Accurate Image Super-Resolution Using Very Deep Convolutional Networks](https://mlsanity.com/paper/1511.04587/) (2015)
- [Holistically-Nested Edge Detection](https://mlsanity.com/paper/1504.06375/) (2015)
- [Universal Style Transfer via Feature Transforms](https://mlsanity.com/paper/1705.08086/) (2017)
- [Controlling Perceptual Factors in Neural Style Transfer](https://mlsanity.com/paper/1611.07865/) (2016)
- [Raiders of the Lost Art](https://mlsanity.com/paper/1909.05677/) (2019)
- [Recovery of underdrawings and ghost-paintings via style transfer by deep convolutional neural networks: A digital tool for art scholars](https://mlsanity.com/paper/2101.10807/) (2021)

## Built on by (most cited first, 1 of 1)

- [Hierarchical Graph-Convolutional Variational AutoEncoding for Generative Modelling of Human Motion](https://mlsanity.com/paper/2111.12602/) (2021)

## Abstract (arXiv)

We apply generative adversarial convolutional neural networks to the problem of style transfer to underdrawings and ghost-images in x-rays of fine art paintings with a special focus on enhancing their spatial resolution. We build upon a neural architecture developed for the related problem of synthesizing high-resolution photo-realistic image from semantic label maps. Our neural architecture achieves high resolution through a hierarchy of generators and discriminator sub-networks, working throughout a range of spatial resolutions. This coarse-to-fine generator architecture can increase the effective resolution by a factor of eight in each spatial direction, or an overall increase in number of pixels by a factor of 64. We also show that even just a few examples of human-generated image segmentations can greatly improve -- qualitatively and quantitatively -- the generated images. We demonstrate our method on works such as Leonardo's Madonna of the carnation and the underdrawing in his Virgin of the rocks, which pose several special problems in style transfer, including the paucity of representative works from which to learn and transfer style information.