| Summary: | This thesis presents a detailed study on psychoacoustic modeling for feature extraction for robust speech recognition. In an automatic speech recognition (ASR) system, feature extraction is critical to determining the recognizer's performance. The most popular feature vectors for ASR are Mel Frequency Cepstral Coefficients (MFCC). However, it is also well known that its performance drops dramatically under noisy condition. One of the objectives of this thesis is to improve the robustness of a recognizer. Compared to an ASR system, human is good at tolerating background noise, hence psychoacoustic modeling of human hearing system is investigated and integrated into speech features extraction process of a speech recognizer to increase the robustness of it.
|
|---|