Speaker Recognition using Vector Quantization by Kekre’s Fast Codebook Generation Algorithm

1. System accepts Voice inputs only 2. Database will be containing the stored voice data 3. Feature Extraction of voice input will be performed after providing the input 4. Feature Extraction will be performed for the Voices which are stored in the database as well vector quantization takes place here, and we need to generate a codebook of the vectors 5. Training data will be used as well 6. Pattern Matching between the feature vector of the input voice and codebook will be performed 7. Recognition decision will be taken whether the voice of the speaker is recognized or not 8. Recognition output will be given