Abstract
The importance of video meetings in today’s fast-paced and interconnected world cannot be overstated. With the rise of remote work and virtual collaboration, video meetings provide a cost-effective, efficient, and convenient way to communicate and collaborate with colleagues, clients, and partners from anywhere in the world. They have also transformed the way businesses provide customer service by offering a more personalized and engaging experience, providing real-time solutions, and improving customer support. However, the implementation of facial expression recognition (FER) technology in video meetings is still facing significant challenges. The lack of standard datasets for evaluation, and temporal variations of facial expressions are some of the major obstacles that need to be addressed to develop robust FER models. Despite these limitations, the potential applications of FER technology are vast, and it has the potential to revolutionize many aspects of our lives, including improving human-computer interaction, enhancing security and surveillance. This study investigates a research problem that aims to design and develop a novel methodology for FER in videos using transformer models to detect customer satisfaction during video meetings. With this study, we aim to contribute to the development of a software architecture that uses three different pre-trained transformer models for FER.
Keywords
Subject Areas
Citations by Year
Related YTU Activities
Sustainability activities and projects connected to this article.
