A new pitch determination algorithm in frequency domain

This paper proposes a novel pitch determination algorithm with binary lateral inhibition network (BLIN). The procedure of this algorithm is shown as follows. A thread-like spectrum, which is composed of the fundamental frequency and its harmonic components, is acquired by applying a BLIN to the short time spectrum of the speech signal. Then the pitch is determined by the average interval of harmonics, which is searched by a flexible strategy. The algorithm is evaluated on COSDIC speech database. For the purpose of comparison, the same results of the cepstrum-and antocorrelaticn-based pitch determination algorithm are also presented. The results show that the new algorithm is more accurate and more robust to white noise than the other two algorithms.