作者: Xinbo Gao , Nannan Wang , Dacheng Tao , Xuelong Li
DOI: 10.1109/TCSVT.2012.2198090
关键词: Computer science 、 Image retrieval 、 Artificial intelligence 、 Feature extraction 、 Pattern recognition 、 Sketch 、 Computer vision 、 Face (geometry) 、 Sparse approximation 、 Image (mathematics) 、 Visual Word 、 Facial recognition system
摘要: Sketch-photo synthesis plays an important role in sketch-based face photo retrieval and photo-based sketch systems. In this paper, we propose automatic sketch-photo algorithm based on sparse representation. The proposed method works at patch level is composed of two steps: neighbor selection (SNS) for initial estimate the pseudoimage (pseudosketch or pseudophoto) sparse-representation-based enhancement (SRE) further improving quality synthesized image. SNS can find closely related neighbors adaptively then generate pseudoimage. SRE, a coupled representation model first constructed to learn mapping between patches patches, patch-derivative-based subsequently applied enhance photos sketches. Finally, four modes, namely, sketch-based, photo-based, pseudosketch-based, pseudophoto-based are proposed, developed by using Extensive experimental results illustrate effectiveness algorithms.