Speech Enhancement Using Multiple Bases Realizations

点击翻译按钮获取中文摘要

9308925 DeBrunner The goal of this project is to investigate the use of multiple bases representations of digitized speech signals to improve the performance of speech enhancement and speech syntheses systems. In the past it has been shown that the use of several basis sets simultaneously provides better energy packing than a single basis set for speech compression systems. This research will determine the quality of reconstructed speech using multiple bases decompositions with the overall objective of improving the performance of digital hearing aids. The proposed technique will be compared to sinusoidal model based algorithms and traditional noise suppression and cancellation techniques. This research is important since it has the potential to advance the fields of speech enhancement, speech synthesis and speech recognition as well as provide useful results that could be used to design new digital hearing aids. ***