Flexible Example-Based Image Enhancement with Task Adaptive Global Feature Self-Guided Network

Dario Kneubuehler, Shuhang Gu, Luc Van Gool, Radu Timofte

ECCVW 2020 pp. 343-358

doi:10.1007/978-3-030-67070-2_21 /eccvw/2020/kneubuehler2020eccvw-flexible/

Abstract

We propose the first practical multitask image enhancement network, that is able to learn one-to-many and many-to-one image mappings. We show that our model outperforms the current state of the art in learning a single enhancement mapping, while having significantly fewer parameters than its competitors. Furthermore, the model achieves even higher performance on learning multiple mappings simultaneously, by taking advantage of shared representations. Our network is based on the recently proposed SGN architecture, with modifications targeted at incorporating global features and style adaption. Finally, we present an unpaired learning method for multitask image enhancement, that is based on generative adversarial networks (GANs).

PDF ECCVW Semantic Scholar

Cite

Text

Kneubuehler et al. "Flexible Example-Based Image Enhancement with Task Adaptive Global Feature Self-Guided Network." European Conference on Computer Vision Workshops, 2020. doi:10.1007/978-3-030-67070-2_21

Markdown

[Kneubuehler et al. "Flexible Example-Based Image Enhancement with Task Adaptive Global Feature Self-Guided Network." European Conference on Computer Vision Workshops, 2020.](https://mlanthology.org/eccvw/2020/kneubuehler2020eccvw-flexible/) doi:10.1007/978-3-030-67070-2_21

BibTeX

@inproceedings{kneubuehler2020eccvw-flexible,
  title     = {{Flexible Example-Based Image Enhancement with Task Adaptive Global Feature Self-Guided Network}},
  author    = {Kneubuehler, Dario and Gu, Shuhang and Van Gool, Luc and Timofte, Radu},
  booktitle = {European Conference on Computer Vision Workshops},
  year      = {2020},
  pages     = {343-358},
  doi       = {10.1007/978-3-030-67070-2_21},
  url       = {https://mlanthology.org/eccvw/2020/kneubuehler2020eccvw-flexible/}
}