The use of deep learning and mean shift to learn global and local processing in human visual perception
Wei-Wen Hsu, Min Zhang, Chung‐Hao Chen, Wen-Chao Yang · 2016
The purpose of this research is to provide a puzzle-based framework to study how global and local information is interacted on human's visual perception during the decision making process. Since the Deep Convolutional Neural Networks (DCNN) has shown the state of the art performance in image classification and object detection, DCNN can output scores to reflect the level of global information, which is highly similar to human's recognition for partial objects in images, i.e. puzzle pieces. Besides, the local information is also taken into consideration in decision making when playing puzzle, therefore, the puzzle-based instrument is proposed in this paper to study how local and global information can be integrated in decision making. The results suggest the proposed method in evaluations of local and global aspects can reflect subjects' behaviors in decision making successfully.