A prototype of
The overall logic of AVET is to first expose the user to a Virtual Environment (VE) to finish a task (e.g., listen to a virtual teacher’s talking) with potential distractions. At the second step, the system will detect the user’s attention with eye-tracking data and physiological signals, then remove all the detected distractions (e.g., a red mug, the virtual teacher’s necklace) from the VE. The last step is to gradually put all distractions back into the VE, which will allow the user exposure to one more distraction object at a time. In this step, the system will provide positive/negative reinforcements to reinforce the desired specific attention. Figure 1 – a) shows the initial VE which is a classroom and a virtual teacher stands in front of the user. The Cyan and the red balls representing the current left and right eye gazing points of the user. The eye gazing point in this figure is just for demonstration purpose, it is hidden from the actual users. b) shows the user is gazing at the teacher. c) shows the user is distracted by the globe and gazed at the globe. d) shows another distraction – the table. e) shows that all the distractions have been removed from the scene. f) shows one of the

We conducted a preliminary evaluation