Providing a single ground-truth for illuminant estimation for the ColorChecker dataset

Ghalia Hemrit; Graham David Finlayson; Arjan Gijsenij; Peter Vincent Gehler; Simone Bianco; Mark Drew; Brian Funt; Lilong Shi

doi:10.1109/TPAMI.2019.2919824

Providing a single ground-truth for illuminant estimation for the ColorChecker dataset

Ghalia Hemrit, Graham David Finlayson, Arjan Gijsenij, Peter Vincent Gehler, Simone Bianco, Mark Drew, Brian Funt, Lilong Shi

Research output: Contribution to journal › Article › peer-review

14 Citations (Scopus)

52 Downloads (Pure)

Abstract

The ColorChecker dataset is one of the most widely used image sets for evaluating and ranking illuminant estimation algorithms. However, this single set of images has at least 3 different sets of ground-truth (i.e. correct answers) associated with it. In the literature it is often asserted that one algorithm is better than another when the algorithms in question have been tuned and tested with the different ground-truths. In this short correspondence we present some of the background as to why the 3 existing ground-truths are different and go on to make a new single and recommended set of correct answers. Experiments reinforce the importance of this work in that we show that the total ordering of a set of algorithms may be reversed depending on whether we use the new or legacy ground-truth data.

Original language	English
Pages (from-to)	1-3
Number of pages	3
Journal	IEEE Transactions on Pattern Analysis and Machine Intelligence
Volume	14
Issue number	8
DOIs	https://doi.org/10.1109/TPAMI.2019.2919824
Publication status	Published - 1 Jul 2019

Access to Document

10.1109/TPAMI.2019.2919824Licence: CC BY

Published_VersionFinal published version, 892 KBLicence: CC BY

https://ieeexplore.ieee.org/document/8751985/Licence: CC BY

Graham Finlayson
- G.Finlaysonuea.acuk
- School of Computing Sciences - Professor of Computing Science
- Colour and Imaging Lab - Member
Person: Research Group Member, Academic, Teaching & Research

A Spatio-chromatic colour appearance model for retargeting high dynamic range image appearance across viewing conditions
Finlayson, G.
Engineering and Physical Sciences Research Council
1/02/17 → 31/01/20
Project: Research
Colour space homography
Finlayson, G. & Trollope, P.
Engineering and Physical Sciences Research Council
28/02/15 → 27/02/19
Project: Research

Cite this

@article{e7a3176399814d7d8d662f09a82462df,

title = "Providing a single ground-truth for illuminant estimation for the ColorChecker dataset",

abstract = "The ColorChecker dataset is one of the most widely used image sets for evaluating and ranking illuminant estimation algorithms. However, this single set of images has at least 3 different sets of ground-truth (i.e. correct answers) associated with it. In the literature it is often asserted that one algorithm is better than another when the algorithms in question have been tuned and tested with the different ground-truths. In this short correspondence we present some of the background as to why the 3 existing ground-truths are different and go on to make a new single and recommended set of correct answers. Experiments reinforce the importance of this work in that we show that the total ordering of a set of algorithms may be reversed depending on whether we use the new or legacy ground-truth data.",

author = "Ghalia Hemrit and Finlayson, {Graham David} and Arjan Gijsenij and Gehler, {Peter Vincent} and Simone Bianco and Mark Drew and Brian Funt and Lilong Shi",

year = "2019",

month = jul,

day = "1",

doi = "10.1109/TPAMI.2019.2919824",

language = "English",

volume = "14",

pages = "1--3",

journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",

issn = "0162-8828",

publisher = "The Institute of Electrical and Electronics Engineers (IEEE)",

number = "8",

}

TY - JOUR

T1 - Providing a single ground-truth for illuminant estimation for the ColorChecker dataset

AU - Hemrit, Ghalia

AU - Finlayson, Graham David

AU - Gijsenij, Arjan

AU - Gehler, Peter Vincent

AU - Bianco, Simone

AU - Drew, Mark

AU - Funt, Brian

AU - Shi, Lilong

PY - 2019/7/1

Y1 - 2019/7/1

N2 - The ColorChecker dataset is one of the most widely used image sets for evaluating and ranking illuminant estimation algorithms. However, this single set of images has at least 3 different sets of ground-truth (i.e. correct answers) associated with it. In the literature it is often asserted that one algorithm is better than another when the algorithms in question have been tuned and tested with the different ground-truths. In this short correspondence we present some of the background as to why the 3 existing ground-truths are different and go on to make a new single and recommended set of correct answers. Experiments reinforce the importance of this work in that we show that the total ordering of a set of algorithms may be reversed depending on whether we use the new or legacy ground-truth data.

AB - The ColorChecker dataset is one of the most widely used image sets for evaluating and ranking illuminant estimation algorithms. However, this single set of images has at least 3 different sets of ground-truth (i.e. correct answers) associated with it. In the literature it is often asserted that one algorithm is better than another when the algorithms in question have been tuned and tested with the different ground-truths. In this short correspondence we present some of the background as to why the 3 existing ground-truths are different and go on to make a new single and recommended set of correct answers. Experiments reinforce the importance of this work in that we show that the total ordering of a set of algorithms may be reversed depending on whether we use the new or legacy ground-truth data.

U2 - 10.1109/TPAMI.2019.2919824

DO - 10.1109/TPAMI.2019.2919824

M3 - Article

VL - 14

SP - 1

EP - 3

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

SN - 0162-8828

IS - 8

ER -

Providing a single ground-truth for illuminant estimation for the ColorChecker dataset

Abstract

Access to Document

Profiles

Graham Finlayson

Projects

A Spatio-chromatic colour appearance model for retargeting high dynamic range image appearance across viewing conditions

Colour space homography

Cite this