English

Fast Fourier Color Constancy

Computer Vision and Pattern Recognition 2020-08-14 v3

Abstract

We present Fast Fourier Color Constancy (FFCC), a color constancy algorithm which solves illuminant estimation by reducing it to a spatial localization task on a torus. By operating in the frequency domain, FFCC produces lower error rates than the previous state-of-the-art by 13-20% while being 250-3000 times faster. This unconventional approach introduces challenges regarding aliasing, directional statistics, and preconditioning, which we address. By producing a complete posterior distribution over illuminants instead of a single illuminant estimate, FFCC enables better training techniques, an effective temporal smoothing technique, and richer methods for error analysis. Our implementation of FFCC runs at ~700 frames per second on a mobile device, allowing it to be used as an accurate, real-time, temporally-coherent automatic white balance algorithm.

Keywords

Cite

@article{arxiv.1611.07596,
  title  = {Fast Fourier Color Constancy},
  author = {Jonathan T. Barron and Yun-Ta Tsai},
  journal= {arXiv preprint arXiv:1611.07596},
  year   = {2020}
}

Comments

CVPR 2017

R2 v1 2026-06-22T17:01:41.485Z