Temporal interpolation of adjacent spectra

摘要

The present invention generally relates to speech enhancement technology applied in various applications such as hands-free telephone systems, speech dialog systems, or in-car communication systems. The basic idea of this new invention is to exploit the redundancy of succeeding FFT spectra and use this for computing interpolated temporal supporting points. Using this interpolated subband signals the aliasing properties of the reference subband signals can be improved and the performance of echo cancellation filters can be improved drastically. As a result a good echo performance can be achieved even at high sub-sampling rates. Due the larger sub-sampling rate the computational complexity of the overall speech enhancement system can be reduced by a large amount (30 to 50 percent).