Transductive Face Sketch-Photo Synthesis

In this paper, we presented a transductive face sketchphoto synthesis method. We first constructed a probabilistic framework to model the sketch-photo generation process. We then gave an alternative optimization method to solve the proposed problem, which can be further decomposed into a number of subproblems. Experiment results showed that the proposed method can achieve a better visual image quality than several other…

Details (available soon)

Multiple Representations-Based Face Sketch-Photo Synthesis

In this paper, we proposed a novel MrFSPS method. Unlike existing methods, our method adaptively combines multiple representations to represent an image patch. To effectively and efficiently combine these features for face sketch–photo synthesis, we developed a Markov networks-based framework to adaptively learn the weights of…

Details