Gradient-descent based window optimization for linear prediction analysis
Wai C. Chu · 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03). · 2003
The autocorrelation method of linear prediction (LP) analysis relies on a window for data extraction; we propose an approach to optimize the window based on gradient-descent. It is shown that the optimized window has improved performance with respect to popular windows, such as Hamming. The technique has potential in quality improvement for many LP-based speech coders.