The main goal of this paper is to present an adaptive filter system using NLMS (normalized least mean square) algorithm for noise cancellation. The proposed algorithm has less computational complexity and better convergence property than the former algorithms like spectral subtraction algorithm. We use TIMIT criterion voice and Noisex-92 for the experiment. The experimental result shows the feasibility of our algorithm can filter noise from voice effectively.

NLMS algorithm, noise cancellation, speech processing.