USING DEEP LEARNING FOR THE BRIGHTNESS DETECTION OF VR SCENES ON THE PREPARATION OF USABILITY TESTING
Abstract - The purpose of this study is a prior preparation using deep learning for an implementation of a planned usability testing on the subject of the brightness of VR contents. Images of pupil dilations are collected reflecting controlled levels of brightness in the immersed VR environment. Those image data are used in deep learning, the convolutional neural network, to train and test the model to distinguish from the pupil dilation patterns under different levels of brightness. Then the trained model will be used in the usability testing of VR developments. The scope of this paper is to demonstrate the acquisition of image data and data processing for training and testing.
Keywords - Deep Learning, VR Content Brightness, Usability Testing.