论文信息 - Design of BP Speaker Recognition System Based on KPCA-MFCC Parameter Optimization

Design of BP Speaker Recognition System Based on KPCA-MFCC Parameter Optimization

The recognition of the speaker through machine learning algorithm has become a hot spot of research. On the basis of speaker recognition based on BP and traditional MFCC characteristic parameters, the feature parameters of MFCC are reduced by KPCA algorithm, and the BP neural network algorithm is used as the back-end recognition model to classify the speaker. The improved algorithm is simulated on the MATLAB platform and compared with the traditional PCA algorithm. The experimental results show that the improved algorithm has a great improvement in recognition efficiency and recognition accuracy and has a good research value.

[1] Tushar Ratanpara,et al. Singer Identification Using MFCC and LPC Coefficients from Indian Video Songs , 2015 .

[2] Xu Chen,et al. Application of Improved BP Neural Network in the Frequency Identification of Piano Tone , 2014 .

[3] Wei Gong,et al. A Research of Speaker Recognition Based on VQ and MFCC , 2014 .

[4] Weiping Hu,et al. Optimization of Pathological Voice Feature Based on KPCA and SVM , 2014, CCBR.

[5] Faten Fayala,et al. A principal component analysis (PCA) method for predicting the correlation between some fabric parameters and the drape , 2014 .