Gait recognition based on DWT and t-SNE

In order to improve the recognition performance and solve the problem of computational complexity caused by the high-dimensional data in human identification, a gait recognition method based on manifold learning is proposed in this paper. Firstly, gait energy image (GEI) of a walking person is abstracted from a gait image sequence. And then discrete wavelet decomposition (DWT) and t-Distributed Stochastic Neighbor Embedding (t-SNE) method is applied to reduce the dimension of high-dimensional GEI data. Finally the support vector machine (SVM) models are trained by the decomposed feature vectors, and the gaits are classified by the trained SVM models at last. Experimental results show that the proposed feature extraction method is efficient in reducing computational complexity and preserving image information.