Speaker Recognition Based on i-vector and Improved Local Preserving Projection

In this paper,a improved local preserve projection algorithm is proposed in order to enhance the recognition performance of the i-vector speaker recognition system under unpredicted noise environment. First , the non zero eigenvalue is rejected when we solve the optimal objective function and only the value greater than zero are used. A mapping matrix is obtained by solving a generalized eigenvalue problem, so can settle the singular value problem always occurred in traditional local preserve projection algorithm. The experiment results shown that the recognition performance of the method proposed in this paper is improved under several kinds of noise environments.