Face recognition method based on robust adaptive graph structure learning algorithm

Abstract

The invention discloses a face recognition method based on a robust adaptive graph structure learning algorithm. The robust adaptive graph structure learning algorithm is applied to a face recognition task, and the robust adaptive graph structure learning algorithm is different from an existing graph learning method, global structure information and local structure information of data are mined by utilizing a self-representation and self-adaptive neighbor method. Meanwhile, in order to reduce the influence of noise points on a graph structure, the robustness of the algorithm to noise is improved by introducing manifold constraints based on l2,1 norms, so that the purpose of robust graph construction is achieved. According to the method, local structure information and global structure information of data are fully mined, and the effectiveness of a graph structure is improved; meanwhile, manifold constraints based on l2,1 norms are introduced into the graph learning process, the robustness of graph learning is enhanced, and the recognition effect is improved.

Description

technical field [0001] The disclosure of the present invention relates to the technical field of machine learning, in particular to a face recognition method based on a robust adaptive graph structure learning algorithm. Background technique [0002] As a branch of biometrics, face recognition technology has been widely used in the fields of biology, human-computer interaction and information security, and it has become a very important research topic in the fields of pattern recognition and computer vision. Although a large number of researchers have proposed various related algorithms to improve the performance of face recognition systems, it is still a very challenging problem. This is because face images captured in real environments are extremely susceptible to effects such as lighting, age, pose, facial expression, and camouflage. Moreover, factors such as occlusion and noise will also affect the performance of face recognition algorithms. If the influence of these f...

AI Technical Summary

Problems solved by technology

Although LCLSR considers the local structural relationship of the data, there are still the following limitations: On the one hand, the objective function of LCLSR is based on l 2 Norm, so the method is sensitive to noise; on the other hand, the sample reconstruction process ignores the relationship between sample representations, that is, similar samples should have similar coded representation coefficients, thus weakening the effectiveness of graph learning

Images