Adaptive representation-based face sketch-photo synthesis

Face sketch synthesis plays an important role in public security and digital entertainment. Most existing face sketch synthesis methods employ pixel intensities or other features to synthesize the whole image. Since different regions have their distinctive properties, they should be represented by different features. This paper presents a novel adaptive representation-based face sketch synthesis method that different regions are represented with different features. It combines multiple features generated from face images pass through several filters and deploys Markov networks to exploit the interacting relationships between neighboring image patches. The proposed model is optimized using an alternative optimization strategy. Experimental results on the Chinese University of Hong Kong (CUHK) face sketch database (CUFS) demonstrate the effectiveness of the proposed method in comparison to state-of-the-art methods.