Abstract
The objective of this paper is to argue that data mattered more than network in terms of color constancy. Computational color constancy is a linear operation device-dependent, which is part of the camera imaging pipeline. We extend the dataset based on this pipeline and prove that the scene illumination can be predicted using a very simple network as long as the dataset is large enough and evenly distributed. In the process of expanding the dataset, firstly, we remove illumination color casts in images which is ground-truth illumination color and then casts randomly generated evenly distributed illumination colors in images. We randomly generate five labels for each image and then work on the image to obtain this dataset. Using this dataset, we introduce a very simple network that is able to compute the color mapping function to correct the image’s colors. Experiments on our new datasets demonstrate that the method of this paper significantly outperforms the state-of-the-art color constancy methods.
Similar content being viewed by others
References
Bianco S, Cusano C. Quasi-unsupervised color constancy [C]. Proceedings of the IEEE conference on computer vision and pattern recognition. 2019: 12212–12221.
Laakom F, Raitoharju J, Iosifidis A, et al. Color constancy convolutional autoencoder [J]. arXiv preprint arXiv:1906.01340, 2019.
Sidorov O. Conditional GANs for multi-illuminant color constancy: revolution or yet another approach? [C]. Proceedings of the IEEE conference on computer vision and pattern recognition workshops. 2019: 0–0.
Afifi M. Semantic white balance: Semantic color constancy using convolutional neural network [J]. arXiv preprint arXiv:1802.00153, 2018.
Afifi M, Brown M S. Sensor-independent illumination estimation for DNN models [J]. arXiv preprint arXiv:1912.06888, 2019.
Buchsbaum, G.: A spatial processor model for object colour perception [J]. J. Franklin Inst. 310(1), 1–26 (1980)
Van De Weijer, J., Gevers, T., Gijsenij, A.: Edge-based color constancy [J]. IEEE Trans. Image Process. 16(9), 2207–2214 (2007)
Afifi, M., Punnappurath, A., Finlayson, G., et al.: As-projective-as-possible bias correction for illumination estimation algorithms [J]. JOSA A 36(1), 71–78 (2019)
Shi W, Loy C C, Tang X. Deep specialized network for illuminant estimation [C]. European conference on computer vision. Springer, Cham, 2016: 371–387.
Bianco, S., Cusano, C., Schettini, R.: Single and multiple illuminant estimation using convolutional neural networks [J]. IEEE Trans. Image Process. 26(9), 4347–4362 (2017)
Bianco S, Cusano C, Schettini R. Color constancy using CNNs[C]. Proceedings of the IEEE conference on computer vision and pattern recognition workshops. 2015: 81–89.
Hu Y, Wang B, Lin S. Fc4: Fully convolutional color constancy with confidence-weighted pooling[C]. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 4085–4094.
Li, X., Zhu, Y., Han, J., et al.: TDCC: top-down semantic aggregation for colour constancy [J]. IET Image Proc. 13(11), 1944–1950 (2019)
Afifi M, Price B, Cohen S, et al. When color constancy goes wrong: correcting improperly white-balanced images [C]. Proceedings of the IEEE conference on computer vision and pattern recognition. 2019: 1535–1544.
Laakom, F., Passalis, N., Raitoharju, J., et al.: Bag of color features for color constancy [J]. IEEE Trans. Image Process. 29, 7722–7734 (2020)
Barron J T, Tsai Y T. Fast fourier color constancy[C]. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 886–894.
Laakom F, Raitoharju J, Iosifidis A, et al. Probabilistic color constancy [J]. arXiv preprint arXiv:2005.02730, 2020.
Qian Y, Chen K, Nikkanen J, et al. Recurrent color constancy[C]. Proceedings of the IEEE international conference on computer vision. 2017: 5458–5466.
Banić N, Koščević K, Lončarić S. Unsupervised learning for color constancy [J]. arXiv preprint arXiv:1712.00436, 2017.
Finlayson, G.D., Zakizadeh, R.: Reproduction angular error: an improved performance metric for illuminant estimation [J]. Perception 310(1), 1–26 (2014)
Banić N, Koščević K, Subašić M, et al. Crop: Color constancy benchmark dataset generator [J]. arXiv preprint arXiv:1903.12581, 2019.
Gong H. Convolutional mean: A simple convolutional neural network for illuminant estimation [J]. arXiv preprint arXiv:2001.04911, 2020.
Hernandez-Juarez D, Parisot S, Busam B, et al. A multi-hypothesis approach to color constancy [C]. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 2270–2280.
Afifi M, Brown M S. What else can fool deep learning? Addressing color constancy errors on deep neural network performance [C]. Proceedings of the IEEE international conference on computer vision. 2019: 243–252.
Afifi M, Brown M S. Deep White-Balance Editing [C].Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition. 2020: 1397–1406.
Gross, D., Audenaert, K., Eisert, J.: Evenly distributed unitaries: on the structure of unitary designs [J]. J. Math. Phys. 48(5), 052104 (2007)
Springenberg J T, Dosovitskiy A, Brox T, et al. Striving for simplicity: The all convolutional net [J]. arXiv preprint arXiv:1412.6806, 2014.
Cheng Z, Yang Q, Sheng B. Deep colorization[C].Proceedings of the IEEE International conference on computer vision. 2015: 415–423.
Land, E.H., McCann, J.J.: Lightness and retinex theory [J]. Josa 61(1), 1–11 (1971)
Barnard, K., Martin, L., Coath, A., et al.: A comparison of computational color constancy algorithms. Ii. experiments with image data [J]. IEEE Trans. Image Proces. 11(9), 985–996 (2002)
Chakrabarti, A., Hirakawa, K., Zickler, T.: Color constancy with spatio-spectral statistics [J]. IEEE Trans. Pattern Anal. Mach. Intell. 34(8), 1509–1519 (2011)
Finlayson, G.D., Trezzi, E.: Shades of gray and colour constancy [C] color and imaging conference. Soc. Imag. Sci. Technol. 2004(1), 37–41 (2004)
Joze, H.R.V., Drew, M.S., Finlayson, G.D., et al.: The role of bright pixels in illumination estimation[C] color and imaging conference. Soc. Imag. Sci. Technol. 2012(1), 41–46 (2012)
Cheng D, Price B, Cohen S, et al. Effective learning-based illuminant estimation using simple features[C].Proceedings of the IEEE conference on computer vision and pattern recognition. 2015: 1000–1008.
Gehler P V, Rother C, Blake A, et al. Bayesian color constancy revisited[C].2008 IEEE Conference on computer vision and pattern recognition. IEEE, 2008: 1–8.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Du, ZM., Li, HA. & Fan, XY. In color constancy: data mattered more than network. Machine Vision and Applications 32, 61 (2021). https://doi.org/10.1007/s00138-021-01190-w
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00138-021-01190-w