101. Learning Illuminant Estimation from Object Recognition
- Author
-
Marco Buzzelli, Joost van de Weijer, Raimondo Schettini, Buzzelli, M, van de Weijer, J, and Schettini, R
- Subjects
FOS: Computer and information sciences ,Ground truth ,Color constancy ,Computer science ,business.industry ,Computer Vision and Pattern Recognition (cs.CV) ,Deep learning ,Cognitive neuroscience of visual object recognition ,ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION ,Computer Science - Computer Vision and Pattern Recognition ,INF/01 - INFORMATICA ,Pattern recognition ,Standard illuminant ,02 engineering and technology ,01 natural sciences ,Task (project management) ,010309 optics ,0103 physical sciences ,0202 electrical engineering, electronic engineering, information engineering ,Illuminant estimation, computational color constancy, semi-supervised learning, deep learning, convolutional neural networks ,020201 artificial intelligence & image processing ,Artificial intelligence ,business - Abstract
In this paper we present a deep learning method to estimate the illuminant of an image. Our model is not trained with illuminant annotations, but with the objective of improving performance on an auxiliary task such as object recognition. To the best of our knowledge, this is the first example of a deep learning architecture for illuminant estimation that is trained without ground truth illuminants. We evaluate our solution on standard datasets for color constancy, and compare it with state of the art methods. Our proposal is shown to outperform most deep learning methods in a cross-dataset evaluation setup, and to present competitive results in a comparison with parametric solutions., Accepted at ICIP 2018
- Published
- 2018