Design and implementation of the image-based group emotion detector

摘要

Group emotion refers to the emotions of a group of people in a specified space and time interval. It is the result of the interactions among the people and the environment. Group emotion is the group's dynamic processes, and can be seen as the sum of the individuals' emotional states. In practice, the group emotion detection has various applications, such as e-learning systems and video conferencing. In this paper, we proposed a the group emotion detector is based on the multi-level smile detection algorithm. According to the bottom-up view point, the group emotion is defined as the average of the smile degrees on this group of people. The whole system is implemented on the Raspberry Pi board for the compactness and mobility. The implementation shows this approach is with availability in practice. The experimental basically result shows the effectiveness of the Group Emotion Detector.