Fast Fourier Color Constancy (2016-11-23T00:00:00.000000Z)

TL;DR

Fast Fourier Color Constancy (FFCC), a color constancy algorithm which solves illuminant estimation by reducing it to a spatial localization task on a torus, enables better training techniques, an effective temporal smoothing technique, and richer methods for error analysis.

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.

Authors

J. Barron

14 papers

Yun-Ta Tsai

1 papers

TL;DR

Abstract

Authors

References41 items

Burst photography for high dynamic range and low-light imaging on mobile cameras

Deep Specialized Network for Illuminant Estimation

Convolutional Color Constancy

Effective learning-based illuminant estimation using simple features

Efficient illuminant estimation for color constancy using grey pixels

Color Constancy by Learning to Predict Chromaticity from Luminance

Color constancy using CNNs

Efficient Color Constancy with Local Surface Reflectance Statistics

Illuminant estimation for color constancy: why spatial-domain methods work and the role of the color distribution.

Learning Fine-Grained Image Similarity with Deep Ranking

Corrected-Moment Illuminant Estimation

Decoupling algorithms from schedules for easy optimization of image processing pipelines

Bayesian Methods in Structural Bioinformatics

Computational Color Constancy: Survey and Experiments

Color constancy

Automatic color constancy algorithm selection and combination

Bayesian color constancy revisited

Using High-Level Visual Information for Color Constancy

Edge-Based Color Constancy

Color Constancy using Natural Image Statistics

Shades of Gray and Colour Constancy

Estimating Illumination Chromaticity via Support Vector Regression

A comparison of computational color constancy Algorithms. II. Experiments with image data

Color constancy at a pixel.

The perception of shading and reflectance

Nonlinear total variation based noise removal algorithms

Analysis of the retinex theory of color vision.

A spatial processor model for object colour perception

Statistics of Directional Data

Color Dog - Guiding the Global Illumination Estimation to Better Accuracy

Exemplar-Based Colour Constancy and Multiple Illumination

Generalized Gamut Mapping using Image Derivative Structures for Color Constancy

Color Constancy

RECOVERING INTRINSIC SCENE CHARACTERISTICS FROM IMAGES

Lightness and retinex theory.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 1 Color Constancy with Spatio-Sp

References

Table 1: Performance on the Gehler-Shi dataset

M) FFCC -full, 4 channels 1

Re-processed version of the gehler color constancy dataset of 568 images

Spatio-spectral Statistics [10] 3.06 2.58 2.74 0.87 6.17 2.59 Bright-and-dark Colors PCA [11] 2.93 2.33 2.42 0.78 6.13 2.40 Corrected-Moment

Field of Study

Journal Information

Name

Page

Venue Information

Name

Type

URL

Alternate Names