PROJECT TITLE :

BULDP Biomimetic Uncorrelated Locality Discriminant Projection for Feature Extraction in Face Recognition - 2018

ABSTRACT:

This Project develops a brand new dimensionality reduction technique, named biomimetic uncorrelated locality discriminant projection (BULDP), for face recognition. It's based on unsupervised discriminant projection and two human bionic characteristics: principle of homology continuity and principle of heterogeneous similarity. With these 2 human bionic characteristics, we tend to propose a completely unique adjacency coefficient illustration, which does not solely capture the class information between totally different samples, but also reflects the continuity between similar samples and the similarity between different samples. By applying this new adjacency coefficient into the unsupervised discriminant projection, it will be shown that we tend to will remodel the initial information space into an uncorrelated discriminant subspace. A detailed answer of the proposed BULDP is given primarily based on singular price decomposition. Moreover, we have a tendency to conjointly develop a nonlinear version of our BULDP using kernel functions for nonlinear dimensionality reduction. The performance of the proposed algorithms is evaluated and compared with the state-of-the-art ways on four public benchmarks for face recognition. Experimental results show that the proposed BULDP technique and its nonlinear version achieve abundant competitive recognition performance.