A machine learning based speech enhancement method is proposed to improve the intelligibility of whispered speech. A binary mask estimated by a two-class support vector machine （SVM） classifier is used to synthesize the enhanced whisper. A novel noise robust feature called Gammatone feature cosine coefficients （GFCCs） extracted by an auditory periphery model is derived and used for the binary mask estimation. The intelligibility performance of the proposed method is evaluated and compared with the traditional speech enhancement methods. Objective and subjective evaluation results indicate that the proposed method can effectively improve the intelligibility of whispered speech which is contaminated by noise. Compared with the power subtract algorithm and the log-MMSE algorithm, both of which do not improve the intelligibility in lower signal-to-noise ratio （SNR） environments, the proposed method has good performance in improving the intelligibility of noisy whisper. Additionally, the intelligibility of the enhanced whispered speech using the proposed method also outperforms that of the corresponding unprocessed noisy whispered speech.
A novel emotional speaker recognition system （ESRS） is proposed to compensate for emotion variability. First, the emotion recognition is adopted as a pre-processing part to classify the neutral and emotional speech. Then, the recognized emotion speech is adjusted by prosody modification. Different methods including Gaussian normalization, the Gaussian mixture model （GMM） and support vector regression （SVR） are adopted to define the mapping rules of F0s between emotional and neutral speech, and the average linear ratio is used for the duration modification. Finally, the modified emotional speech is employed for the speaker recognition. The experimental results show that the proposed ESRS can significantly improve the performance of emotional speaker recognition, and the identification rate （IR） is higher than that of the traditional recognition system. The emotional speech with F0 and duration modifications is closer to the neutral one.
A cascaded projection of the Gaussian mixture model algorithm is proposed.First,the marginal distribution of the Gaussian mixture model is computed for different feature dimensions, and a number of sub-classifiers are generated using the marginal distribution model.Each sub-classifier is based on different feature sets.The cascaded structure is adopted to fuse the sub-classifiers dynamically to achieve sample adaptation ability.Secondly,the effectiveness of the proposed algorithm is verified on electrocardiogram emotional signal and speech emotional signal.Emotional data including fidgetiness,happiness and sadness is collected by induction experiments.Finally,the emotion feature extraction method is discussed,including heart rate variability, the chaotic electrocardiogram feature and utterance level static feature.The emotional feature reduction methods are studied, including principle component analysis,sequential forward selection, the Fisher discriminant ratio and maximal information coefficient.The experimental results show that the proposed classification algorithm can effectively improve recognition accuracy in two different scenarios.
