Speech Enhancement by Spectral Component Selection

摘要

Most algorithms for speech enhancement in spectral domain focus on the acquisition of an estimator of the clean speech parameter, such as spectrum or amplitude. Enhanced speech quality with these methods mainly depends upon the accuracy of the estimator. When the signal-to-noise ratio (SNR) becomes lower, e.g., less than 3DB, the enhanced speech often shows unsatisfied quality . Here, we propose a new method for speech enhancement in spectral domain. It bases on our view that speech characteristic is perceived mainly by part of the spectral components which have higher local (or instantaneous) SNRs. These components have a special importance on speech SNR enhancement and speech understand. Identifying these components in the noisy spectrum with a decision system and constructing enhanced speech just by the same nosy amplitude and phase, the result is similar to those popular algorithms. All needed to do is a local SNR decision for each spectral component. The threshold for the decision is derived and two popular methods for local SNR (LSNR) estimation are suggested. It shows better performance under lower SNR signals. Though the number of these kind of spectral components varies with different signal SNR, any algorithms which can pick up more than 90percent of it is sufficient to match any existing methods. Primary considerations and results are shown.