This paper presents a study of driver's voice feature selection and classification for speaker identification in a vehicle security system. The proposed system consisted of a combination of feature extraction using continuous wavelet technique and voice classification using artificial neural network. In the feature extraction, a time-averaged wavelet spectrum based on continuous wavelet transform is proposed. Meanwhile, the artificial neural network techniques were used for classification in the proposed system. In order to verify the effect of the proposed system for classification, a conventional back-propagation neural network (BPNN) and generalized regression neural network (GRNN) were used and compared in the experimental investigation. The experimental results demonstrated the effectiveness of the proposed speaker identification system. The identification rate is about 92% for using BPNN and 97% for using GRNN approach.
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Artificial Intelligence