A novel environment-dependent speech enhancement method with optimized memory footprint

Suhadi Suhadi, Sorel Stan, Tim Fingscheidt

Abstract: Data-driven speech enhancement (Fingscheidt and Suhadi [1]) aims at improving speech quality for voice calls in a specific noise environment. The essence of the method are a set of frequency-dependent weighting rules, indexed by a priori and a posteriori SNRs, which are learned from clean speech and background noise training data. The weighting rules must be stored for each frequency bin separately and take up about 400 kBytes memory, which makes DSP implementations relatively expensive.