Fourier-CPPNs for Image Synthesis

Mattie Tesfaldet, Xavier Snelgrove, David Vázquez

ICCVW 2019 pp. 3173-3176

doi:10.1109/ICCVW.2019.00392 /iccvw/2019/tesfaldet2019iccvw-fouriercppns/

Abstract

Compositional Pattern Producing Networks (CPPNs) are differentiable networks that independently map (x, y) pixel coordinates to (r, g, b) colour values. Recently, CPPNs have been used for creating interesting imagery for creative purposes, e.g., neural art. However their architecture biases generated images to be overly smooth, lacking high-frequency detail. In this work, we extend CPPNs to explicitly model the frequency information for each pixel output, capturing frequencies beyond the DC component. We show that our Fourier-CPPNs (F-CPPNs) provide improved visual detail for image synthesis.

PDF ICCVW Semantic Scholar

Cite

Text

Tesfaldet et al. "Fourier-CPPNs for Image Synthesis." IEEE/CVF International Conference on Computer Vision Workshops, 2019. doi:10.1109/ICCVW.2019.00392

Markdown

[Tesfaldet et al. "Fourier-CPPNs for Image Synthesis." IEEE/CVF International Conference on Computer Vision Workshops, 2019.](https://mlanthology.org/iccvw/2019/tesfaldet2019iccvw-fouriercppns/) doi:10.1109/ICCVW.2019.00392

BibTeX

@inproceedings{tesfaldet2019iccvw-fouriercppns,
  title     = {{Fourier-CPPNs for Image Synthesis}},
  author    = {Tesfaldet, Mattie and Snelgrove, Xavier and Vázquez, David},
  booktitle = {IEEE/CVF International Conference on Computer Vision Workshops},
  year      = {2019},
  pages     = {3173-3176},
  doi       = {10.1109/ICCVW.2019.00392},
  url       = {https://mlanthology.org/iccvw/2019/tesfaldet2019iccvw-fouriercppns/}
}