Image Synthesis for Presenting Facial Information of Learners in e-Learning System

One of the major advantages of recent e-learning systems is that learning-logs are available. By looking at the learninglogs, teachers can obtain learning processes of learners. However, the teachers cannot know facial expressions of the learners and the objects in course materials that the learners focused on while the learners use the e-learning system. In a classroom lecture, it is very important that the teachers can read facial expressions and focus of attention of the learners. This is also the case in asynchronous distance education using e-learning systems. In this paper, we propose storing the facial images of the learners as another kind of learning-log of e-learning system from which the teachers can read facial expressions and focus of attention of the learners. The objects that the learners focus on are the area around the intersections of their lines of sight and the monitor. In order for the teachers to identify the intersection, they need to take the viewpoint that gives the frontal views of the monitor and the learners’ faces at the same time. Moreover, frontal view of the learners’ faces are also helpful for the teachers to understand facial expressions of the learners. However, that viewpoint does not exist in the real world due to the occlusions between the monitor and the learners’ heads. In this paper, we propose obtaining facial images seen through the monitor from the viewpoint behind the monitor by overlaying the images of course material on the frontal facial images of the learners, which are synthesized by using stereo vision and 3D facial models of the learners.