Dictionary-Based Facial Animation Retargeting to Virtual Characters

Abstract

With the rapid growth of the entertainment and media industries, there is more demand for virtual characters that are more realistic and interactive. Facial animation retargeting is a key contributor to making the experience more immersive. This study is dedicated to developing an entirely simple, and fast way of transferring facial animation from video footage to virtual characters. This system is implemented with advanced image processing and a comprehensive dictionary of facial animations to effectively analyze, extract, and replicate key facial expressions from videos. This approach can lower the time and effort needed for manual animation creation while maintaining high fidelity and realism. Furthermore, it supports real-time facial animation retargeting, allowing immediate application in interactive environments. It can be used in various areas like film, video games, advertising and education. Our method is a significant improvement in generating virtual experiences that are more lifelike and engaging to the target user.