An algorithm to estimate pitch frequency under low SNR circumstance is described here.Without consideration of the noise,ordinary pitch estimation algorithms have poor properties in practical circumstance.To solve thi...An algorithm to estimate pitch frequency under low SNR circumstance is described here.Without consideration of the noise,ordinary pitch estimation algorithms have poor properties in practical circumstance.To solve this problem,a new voice noise model is given.By use of the inverse AR parameters filtering and the TLS ESPRIT a new algorithm is proposed which has a good performance under noisy circumstance and a relative low computational complexity.Some of the experimental results are also provided here.展开更多
文摘An algorithm to estimate pitch frequency under low SNR circumstance is described here.Without consideration of the noise,ordinary pitch estimation algorithms have poor properties in practical circumstance.To solve this problem,a new voice noise model is given.By use of the inverse AR parameters filtering and the TLS ESPRIT a new algorithm is proposed which has a good performance under noisy circumstance and a relative low computational complexity.Some of the experimental results are also provided here.
