A new algorithm for end-point detection based on fourth-order cumulants
Ke Chen, Zengfu Wang
Abstract
Ke Chen, Zengfu Wang
Abstract
A new algorithm for end-point detection in noisy background conditions based on the statistical properties of higher-order cumulants of speech signal is presented in this paper. Since the higher-order cumulants are blind for Gaussian signal, the proposed method is especially effective to the problem of end-point detection of speech signals in Gaussian noise. The above conclusions are supported by the experimental results.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A new algorithm for end-point detection in noisy background conditions based on the statistical properties of higher-order cumulants of speech signal is presented in this paper. Since the higher-order cumulants are blind for Gaussian signal, the proposed method is especially effective to the problem of end-point detection of speech signals in Gaussian noise. The above conclusions are supported by the experimental results.
Key concepts: Cumulant, Higher-order statistics, Algorithm, Gaussian, Gaussian noise, Computer science, Detection theory, SIGNAL (programming language)