A GMM-based telephone channel classification for Mandarin speech recognition

The discriminability of different channel modeled by GMM is qualitatively analyzed in this paper. A GMM-based channel classification and speech recognition system in multi-channel environment is proposed. The channel classifier is used to select a most likely HMM from pre-trained HMMs of each specific telephone channel environment. The selected HMM is used as the reference HMM to recognize each utterance. The experimental results show that the proposed system is an efficient framework to enhance the robustness of speech recognition across different channel environment.