Paper Contents
Abstract
Face stylization is a challenging task that requires the transfer of artistic or stylistic attributes while preserving the facial identity. In this work, we propose a novel framework for real-time face stylization using one-shot learning and regression networks. By leveraging the synergy between generic object tracking and regression-based feature transformation, our approach achieves high-quality stylization with minimal inputa single reference image. The method employs a regression network to learn a mapping between the source and target domains, guided by robust object tracking to ensure consistency across frames. This enables seamless stylization in dynamic scenarios, such as video applications, without requiring extensive data or retraining. Experimental results demonstrate that our framework outperforms existing methods in terms of efficiency, adaptability, and visual fidelity, making it a promising solution for real-time face stylization in both artistic and practical applications.
Copyright
Copyright © 2025 Dr.M.Deepa. This is an open access article distributed under the Creative Commons Attribution License.
