CVNov 23, 2016

Fast Fourier Color Constancy

arXiv:1611.07596v3214 citations
Originality Incremental advance
AI Analysis

This provides a real-time, accurate automatic white balance solution for mobile devices, addressing a domain-specific problem with incremental improvements in speed and error reduction.

The paper tackles the problem of illuminant estimation for color constancy by proposing a fast algorithm that reduces it to spatial localization on a torus, achieving 13-20% lower error rates and being 250-3000 times faster than previous state-of-the-art methods.

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.

Code Implementations2 repos
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes