Determining the Appropriate Number of Observers in Eye-Tracking Based Video Saliency Computation
Chuancai Li, Jiayi Xu, Jianjun Li, Xiaoyang Mao · 2018
This paper proposes a new method for suggesting the appropriate number of observers needed in eye-tracking experiments for a given video. We first use the fixation consistency computed as the Similarity between different observers' fixation maps with the ground truth as a feature vector to classify sample video clips into four types. Then these videos are labeled and used to train a support vector machine (SVM) for identifying the class of a given video clip.