唇读中基于像素的特征提取方法的研究

Research of pixel based feature extraction in lip-reading

针对单独视觉通道唇读中的基于像素的特征提取问题,提出一个级联的特征提取策略。首先对图像采用相应的变换,然后对变换结果降维,最后进行特征归一化。基于对几种变换方法的比较与分析,提出利用PCA对DCT和Gabor小波变换结果降维的DCT-PCA和Gabor-PCA方法,与传统人工选择变换系数的方法相比识别率提高了约10%。

This paper concentrates on the pixel based feature extraction in only visual channel lip-reading system.A three-stage cascade visual front end is proposed.The first stage is corresponding transform to be performed over the image,the second stage is to reduce the dimensions of the transformed image,in the third stage all feature vectors are normalized into a uniform scale. We apply PCA to reduce the dimension of DCT and Gabor transformed data called DCT-PCA and Gabor-PCA,which can improve the recognition accuracy by 10% compared with the manually-selected features.