Composite face sketch recognition based on components

Composite face sketch recognition is a challenging problem in face recognition, which is important for law enforcement. Considering composite sketches are generated from facial components, a component-based representation approach was proposed to match composite sketches to photos recently. They extract features from each component and use the fusion of four most discriminative components as the final matching score. However, they ignore the inherent structure of composite sketch and the fusion does not take the full components' information into consideration. This paper presents a novel composite sketch recognition method by extracting two image feature descriptors from components, fusing them at score level, combining facial components with different weights. In addition, we show the detailed procedure of composite sketch generation, which is useful for building related composite database. Experiments on the public composite sketch database (E-PRIP) demonstrate that our proposed method achieves superior performance compared with state-of-the-art methods.