ALCN: Adaptive Local Contrast Normalization
This addresses the challenge of illumination variability in robotics and AR, offering a universal solution that avoids the need for extensive training datasets, though it is incremental as it builds on existing normalization techniques.
The paper tackles the problem of making robotics and augmented reality applications robust to illumination changes by proposing an adaptive illumination normalization method that predicts parameters from input images using a CNN trained jointly with an object recognition network. The method significantly outperforms standard normalization methods, improving robustness in 3D object detection and face recognition without requiring retraining for new applications.
To make Robotics and Augmented Reality applications robust to illumination changes, the current trend is to train a Deep Network with training images captured under many different lighting conditions. Unfortunately, creating such a training set is a very unwieldy and complex task. We therefore propose a novel illumination normalization method that can easily be used for different problems with challenging illumination conditions. Our preliminary experiments show that among current normalization methods, the Difference-of Gaussians method remains a very good baseline, and we introduce a novel illumination normalization model that generalizes it. Our key insight is then that the normalization parameters should depend on the input image, and we aim to train a Convolutional Neural Network to predict these parameters from the input image. This, however, cannot be done in a supervised manner, as the optimal parameters are not known a priori. We thus designed a method to train this network jointly with another network that aims to recognize objects under different illuminations: The latter network performs well when the former network predicts good values for the normalization parameters. We show that our method significantly outperforms standard normalization methods and would also be appear to be universal since it does not have to be re-trained for each new application. Our method improves the robustness to light changes of state-of-the-art 3D object detection and face recognition methods.