A Speech Enhancement Preprocessor for Low Bit Rate Speech Coding

For speech enhancement before low bit rate speech coding in adverse environments, an improved noise estimation approach is proposed to keep speech enhancement performance with lower complexity. First, the noisy speech is transformed into Bark domain and the minima values are tracked. Then, these values are applied to control the smoothing factor of recursive averaging in noise estimator. The proposed approach, based on perceptual model, makes full use of the correlation among frequency bins in the same Bark band to track noise better. Furthermore, the approach has low complexity. The simulation results show that the proposed approach can suppress noise effectively with low speech distortion, and the synthesis speech of 2400 bps MELP (Mixed-Excitation Linear Predictive) vocoder with proposed noise estimation has better speech quality in adverse environments.