Recognizing student facial expressions: A web application

The project described in this paper investigates the idea of performing emotion analysis of a student population participating in active face-to-face classroom instruction. Machine learning algorithms are employed on live recordings collected by webcams that are installed in classrooms. The visualization application required to be remotely accessible by the lecturer so the application was engineered as a web application. The output, being a timeline of student emotions monitored throughout and in parallel with the lecture, serves to enable the lecturer and other interested parties to improve the delivery of education.

I. Introduction

In our modern times massive volumes of data are being collected from multiple varied sources. Sometimes data which are collected for one purpose may as well serve a completely different purpose and thus with better utilization of existing resources a business may enjoy multiple benefits. For example, while most businesses install cameras in their stores to prevent shoplifting further possibilities of processing these camera-captured images are also possible. The images could be analyzed by an intelligent system to reveal customer emotions and potentially determine customer satisfaction, interest and enjoyment while shopping; basically rate the complete client shopping experience! In fact emotions' recognition and analysis may prove to be another powerful leveraging tool for business success. If businesses could sense emotions at all times, they could potentially capitalize on them to sell to the customer at the most favorable moment [1].