Abstract:

Face recognition has been used in security systems, identification verification, terrorist prevention system and clinical image storage. Its related branch facial expression has emerged in use in human machine interface and emotion detection. Usually these projects are at graduate research level because of its complexity and advanced background requirements in high level math and programming skill. However, in order to promote undergraduate research, I present in this paper a research project which undergraduate students can easily follow, implement and develop. This experience is valuable for undergraduate STEM students.

The research project has two parts: face recognition and facial expression. In each part, it is further divided into two subparts: Principle Component Analysis (PCA) algorithm basics and Matlab PCA function programming in face recognition and Viola-Jones algorithm basics and Matlab Viola-Jones function programming in facial expression. In the face recognition part, students will learn the concept of PCA algorithm from my easy to follow notes. They will learn how to extract eigenvectors from face image matrixes. Students will then write codes using popular and easy to use Matlab programming language and PCA functions to identify a target face among a set of image data. In the facial expression part, students will follow my notes to understand how the Viola-Jones algorithm works and how to use it to pin point facial expressions such as sadness or excitement. Students will then write codes using Matlab programming language and Viola-Jones functions to identify a target emotion expression. Once students acquire the basic face recognition and expression detection skill, they can use it to further develop more advance applications such as video game player excitement detection. Through this easy to learn and follow process, undergraduate students can be educated and involved in research and development in the area of face recognition and facial expression detection.

The result of the project was successfully demonstrated. Two students went through the process and at the end they published and presented papers in conferences.

Keywords:

Technology-enhanced learning, undergraduate research, face recognition, facial expression.