Development of facial image processing algorithms for human computer interaction

Abstract

In the era of digital revolution the education and exhibition systems have adopted the advancement of information and communication technology in pursuit of realising targeted content delivery system In museums the interest language and level of understanding of visitors may differ with respect to age race and gender Hence the addressing has to be planned for specific group rather than being for the group of undifferentiated audience Most of the researches in these digitized application focus on developing user friendly interfaces Human Computer Interaction HCI is a key component of the content exchange that can deliver based on the group of targeted audience To achieve these goals the research interest has shifted from generic computer vision systems towards vision modules to solve more specific tasks like detection and classification of people group Vision based HCI uses the human face image for further processing since face offers enormous information about the attributes of a person such as identity ethnicity age gender attractiveness and behaviour However in spite of the advent of novel methodologies such vision based systems are still not ideal compared to human abilities Because the correct rates of these systems strongly depend upon the variations in environmental conditions and face image The objective of this thesis includes the construction of a new transgender face database face detection and classification of demographic attributes like gender age and race from detected face images. newline