Algorithms, Experimentation

Video stylization transfers a source video into an artistic version while maintaining temporal coherence between ad-jacent frames. In this paper, we formulate the unsupervised example-based video stylization with Markov random field model. In our algorithm, we implement an improved optical flow algorithm to maintain temporal coherence while im-prove the accuracy of estimation along motion boundaries. We also extend our algorithm to the application of video personalization, in which human faces keep clear and distin-guishable. A series of techniques are fused in video personal-ization, including face detection and alignment, motion flow, skin detection, and illumination blending. Given a source video and a style template image, our algorithm produces the stylized and/or personalized video(s) automatically. Ex-perimental results demonstrate that our algorithm performs excellently in both video stylization and personalization