两阶段演算法之中文声母语音辨认两阶段演算法之中文声母-德霖.pdfVIP

德霖學報 「第二十二期」 民國 97年 6 月 兩階段演算法之中文聲母語音辨認 A Two-Stage Algorithm for Mandarin Consonant Recognition Ming-Tzau Lin Assistance Professor, Department of Computer and Communication Engineering, De-Lin Institute of Technology Abstract A two-stage algorithm with multiple features for Mandarin consonant recognition was proposed. We classified Mandarin consonant into seven classes, each with similar phonetic and acoustic characteristics. In the first stage of the proposed algorithm, consonant classification was conducted by computing features to construct mixture density models for each feature and for each consonant class, and then we defined for the decision making of consonant classification. Our multi-speaker experiments showed that a high correct percentage (97.9% for the top two selections) could be achieved in our first-stage consonant classification. In the second stage, with the selected top two consonant classes as candidates, detailed consonant recognition is performed by feature extraction of cepstral coefficients followed by the pattern comparison using the continuous hidden Markov model (CHMM) or the segmental probability model (SPM). The experimental results showed that both SPM and CHMM can achieve comparable accuracy percentage of 89.2%, but the SPM requires only about 1/8 processing time. Furthermore, the proposed two-stage algorithm is superior to the one-stage algorithm (omitting the first stage) by about 3% accuracy rate increasing and 12% processing time saving. Keywords: Mandarin consonant recognition, consonant classification, pattern comparison.