In this paper, a manifold subspace learning algorithm based on locality preserving discriminant projection (LPDP) is used for speaker verification. LPDP can overcome the deficiency of the total variability factor anal...In this paper, a manifold subspace learning algorithm based on locality preserving discriminant projection (LPDP) is used for speaker verification. LPDP can overcome the deficiency of the total variability factor analysis and locality preserving projection (LPP). LPDP can effectively use the speaker label information of speech data. Through optimization, LPDP can maintain the inherent manifold local structure of the speech data samples of the same speaker by reducing the distance between them. At the same time, LPDP can enhance the discriminability of the embedding space by expanding the distance between the speech data samples of different speakers. The proposed method is compared with LPP and total variability factor analysis on the NIST SRE 2010 telephone-telephone core condition. The experimental results indicate that the proposed LPDP can overcome the deficiency of LPP and total variability factor analysis and can further improve the system performance.展开更多
Due to the volume conduction,electroencephalogram（EEG） gives a rather blurred image of brain activities. It is a challenge for generating satisfactory performance with EEG. This paper studies the multiple areas fusi...Due to the volume conduction,electroencephalogram（EEG） gives a rather blurred image of brain activities. It is a challenge for generating satisfactory performance with EEG. This paper studies the multiple areas fusion of EEG classifiers to improve the motor imagery EEG classification performance. Two feature extraction methods are employed to extract the feature from three different areas of EEG. One is power spectral density（PSD）, and the other is common spatial patterns（CSP）. Classifiers are designed based on the well-known linear discrimination analysis（LDA）. The fusion of the individual classifiers is realized by means of the Choquet fuzzy integral. It is demonstrated that the proposed method comes with better performance compared with the individual classifier.展开更多
文摘In this paper, a manifold subspace learning algorithm based on locality preserving discriminant projection (LPDP) is used for speaker verification. LPDP can overcome the deficiency of the total variability factor analysis and locality preserving projection (LPP). LPDP can effectively use the speaker label information of speech data. Through optimization, LPDP can maintain the inherent manifold local structure of the speech data samples of the same speaker by reducing the distance between them. At the same time, LPDP can enhance the discriminability of the embedding space by expanding the distance between the speech data samples of different speakers. The proposed method is compared with LPP and total variability factor analysis on the NIST SRE 2010 telephone-telephone core condition. The experimental results indicate that the proposed LPDP can overcome the deficiency of LPP and total variability factor analysis and can further improve the system performance.
文摘Due to the volume conduction,electroencephalogram（EEG） gives a rather blurred image of brain activities. It is a challenge for generating satisfactory performance with EEG. This paper studies the multiple areas fusion of EEG classifiers to improve the motor imagery EEG classification performance. Two feature extraction methods are employed to extract the feature from three different areas of EEG. One is power spectral density（PSD）, and the other is common spatial patterns（CSP）. Classifiers are designed based on the well-known linear discrimination analysis（LDA）. The fusion of the individual classifiers is realized by means of the Choquet fuzzy integral. It is demonstrated that the proposed method comes with better performance compared with the individual classifier.
