Effect of hue shift towards robustness of convolutional neural networks

Abstract

Computer vision systems become deployed in diverse real time systems hence robustness is a major area of concern. As a vast majority of the AI enabled systems are based on convolutional neural networks based models which use 3-channel RGB images as input. It has been shown that the performance of AI systems, such as those used in classification, is impacted by distortions in the images. To date most work has been carried out on distortions such as noise, blur, compression. However, color related changes to images could also impact the performance. Therefore, the goal of this paper is to study the robustness of these models under different hue shifts.

Exploring Effects of Colour and Image Quality in Semantic Segmentation by Deep Learning Methods

“…For image classification, studies have shown [33] that changing the hue angle to red or blue has a significant impact on the performance of deep convolutional networks. Similarly to the previous study, here we shift the hue angle by 60 degrees to create a subset of five classes of images (60,120,180,240,300) as depicted in Figure 3.…”

“…Recent studies [30][31][32] have shown that colour information has a significant impact on image classification tasks. Colour parameters such as the hue angle shift [33] have shown a significant impact on the performance of state-of-theart deep neural networks trained on pristine ImageNet data. Colour information has been exploited successfully in the past by image segmentation algorithms [34][35][36].…”