Robust image classification by coupling low rank and collaborative representation graphs
Junjun Guo, Jing Lei Xin, Shifang Zhang, Qiang Liu · 2017
A single graph cannot comprehensively describe the true relationship among samples with high dimensionality, especially for the graph-based methods which directly calculate the distances among samples in the Euclidean space. To further improve the performance of the method based on the presentation graph, in this paper we propose a new composite graph, called as low rank representation and collaborative representation (LRRCR) graph. The proposed LRRCR graph can obtain more informative knowledge for robust image classification. The experimental results on three real face image data sets show that the proposed method has better performance when compared to the traditional ones.