Iterative Collaborative Representation based Classification for Face Recognition
Wei Huang, Xiaohui Wang, Yinghui Zhu, Jianzhong Li · Signal processing research · 2015
Collaborative representation based classification (CRC) has received much attention in the field of pattern recognition. CRC uses a simple way to code a testing sample as a linear combination of all the training samples and classifies the testing sample into the class with the minimum representation error. But the original algorithm of CRC suffers from the following problem. It must perform the matrix inverse operation, which may cause unstable numerical computation. With this paper, in order to overcome the above problem of the original algorithm of CRC, we propose an iterative collaborative representation based classification (ICRC) algorithm. The experimental results on face recognition show that ICRC not only outperforms CRC but also is able to obtain a much higher accuracy than LRC.