- Developed a system to detect 7 distinct emotions (Anger, Fear, Happiness, Sadness, Disgust, Neutral, Surprise) from Human Speech Samples (using Surrey Audio-Visual Expressed Emotion (SAVEE) Database).
- Concepts utilized: explored first central Spectral Moment time-frequency distribution augmented by low order Cepstral coefficients as features for distinguishing emotions, modeled the extracted features using GMM and adapted it using Maximum Aposteriori (MAP) algorithm from the Universal Background Model(UBM), trained a multi-class SVM classifier with Gaussian kernel for the purpose of classification.
-obtained 12% better accuracy on SAVEE database than the previous best reported systems. In the process of writing a paper on the same
- Keywords: Speech Signal Processing, Emotion Detection, MATLAB, GMM, UBM, SVM
Currently working on the writing part for the project. More details would be uploaded soon once the paper is submitted for review.
- Concepts utilized: explored first central Spectral Moment time-frequency distribution augmented by low order Cepstral coefficients as features for distinguishing emotions, modeled the extracted features using GMM and adapted it using Maximum Aposteriori (MAP) algorithm from the Universal Background Model(UBM), trained a multi-class SVM classifier with Gaussian kernel for the purpose of classification.
-obtained 12% better accuracy on SAVEE database than the previous best reported systems. In the process of writing a paper on the same
- Keywords: Speech Signal Processing, Emotion Detection, MATLAB, GMM, UBM, SVM
Currently working on the writing part for the project. More details would be uploaded soon once the paper is submitted for review.