A new approach for estimating the number of sources that employs independent component analysis (ICA) is discussed. Estimating the number of sources provides information for signal processing applications such as blind source separation (BSS) in the frequency domain. The new method can identify a noise component that includes reverberations by calculating the correlation of the envelopes. The results show that the characteristics of the proposed approach compare with the conventional eigenvalue-based method.
- Convolutive mixtures
- Eigenvalue of correlation matrix
- Independent component analysis
ASJC Scopus subject areas
- Acoustics and Ultrasonics