No abstract

No abstract

In this paper, an efficient algorithm for computing line spectrum pair (LSP) frequencies from linear prediction coefficients is described. Based on the decimation-in-degree algorithm proposed by Wu and Chen, this paper modifies the Ferrari formula to reduce the computational complexity of finding the roots, i.e., LSPs, of the symmetric and anti-symmetric polynomials. By the use of this modified complex-free Ferrari formula, the proposed algorithm is highly accurate and can avoid the operation of complex number. The proposed algorithm provides three different guidelines for computing the 8-, 10-, and 12-order LSP frequencies. It is shown in various experimental results that the proposed algorithm has a considerable computation time improvement when compared with other conventional methods.

a b s t r a c tIn this paper we present a new wavelet-based algorithm for low-cost computation of the cepstrum. It can be used for real time precise pitch determination in automatic speech and speaker recognition systems. Many wavelet families are examined to determine the one that works best. The results confirm the efficacy and accuracy of the proposed technique for pitch extraction.

No abstract