Thermal face recognition using convolutional neural network

With the rapidly development of the face recognition (FR), thermal face recognition has received increasing attention. However, the traditional methods for thermal face recognition mainly concentrate on the hand-crafted feature design, which requires more efforts from human to select and extract features and usually has relatively lower recognition rate. In this paper, we present a convolutional neural network (CNN) architecture for thermal face recognition. CNN is a new type of neural network method which can automatically learn effective features from the raw data. Experiment results on RGB-D-T face database show that our proposed CNN architecture achieves higher recognition rate compared with the traditional recognition such as LBP, HOG and moments invariant.