3D Face Recognition Based on Empirical Mode Decomposition and Sparse Representation
Xing Chen, Yinan Lu, Ran Fang · Proceedings of the 2nd International Conference on Computer Science and Application Engineering · 2018
With1 the research interest increasing in 3D face recognition, many methods for 3D face recognition have emerged in recent years. In this paper, a novel method named as decomposition-based face classification (DFC) is proposed to recognize a 3D face by using empirical mode decomposition (EMD) and sparse representation. A 3D face scan is firstly decomposed into several surfaces by a 3D EMD algorithm, then for each decomposed surface, meshSIFT is used for detecting salient points and extracting the local descriptor, and the dictionary is constructed by concatenating all the local descriptors. Finally, the identification of a probe face can be determined by the multitask sparse representation classification. Experimental results demonstrate that our decomposition-based method can achieve state-of-the-art performance on two benchmark databases, Bosphorus and FRGC2.0.