Evaluation of Voice Acoustics as Predictors of Clinical Depression Scores.

OBJECTIVE: The aim of the present study was to determine if acoustic measures of voice, characterizing specific spectral and timing properties, predict clinical ratings of depression severity measured in a sample of patients using the Hamilton Depression Rating Scale (HAMD) and Beck Depression Inventory (BDI-II). STUDY DESIGN: This is a prospective study. METHODS: Voice samples and clinical depression scores were collected prospectively from consenting adult patients who were referred to psychiatry from the adult emergency department or primary care clinics. The patients were audio-recorded as they read a standardized passage in a nearly closed-room environment. Mean Absolute Error (MAE) between actual and predicted depression scores was used as the primary outcome measure. RESULTS: The average MAE between predicted and actual HAMD scores was approximately two scores for both men and women, and the MAE for the BDI-II scores was approximately one score for men and eight scores for women. Timing features were predictive of HAMD scores in female patients while a combination of timing features and spectral features was predictive of scores in male patients. Timing features were predictive of BDI-II scores in male patients. CONCLUSION: Voice acoustic features extracted from read speech demonstrated variable effectiveness in predicting clinical depression scores in men and women. Voice features were highly predictive of HAMD scores in men and women, and BDI-II scores in men, respectively. The methodology is feasible for diagnostic applications in diverse clinical settings as it can be implemented during a standard clinical interview in a normal closed room and without strict control on the recording environment.

Discriminative common vector method with kernels.

In some pattern recognition tasks, the dimension of the sample space is larger than the number of samples in the training set. This is known as the "small sample size problem". Linear discriminant analysis (LDA) techniques cannot be applied directly to the small sample size case. The small sample size problem is also encountered when kernel approaches are used for recognition. In this paper, we attempt to answer the question of "How should one choose the optimal projection vectors for feature extraction in the small sample size case?" Based on our findings, we propose a new method called the kernel discriminative common vector method. In this method, we first nonlinearly map the original input space to an implicit higher dimensional feature space, in which the data are hoped to be linearly separable. Then, the optimal projection vectors are computed in this transformed space. The proposed method yields an optimal solution for maximizing a modified Fisher's linear discriminant criterion, discussed in the paper. Thus, under certain conditions, a 100% recognition rate is guaranteed for the training set samples. Experiments on test data also show that, in many situations, the generalization performance of the proposed method compares favorably with other kernel approaches.

Discriminative common vectors for face recognition.

In face recognition tasks, the dimension of the sample space is typically larger than the number of the samples in the training set. As a consequence, the within-class scatter matrix is singular and the Linear Discriminant Analysis (LDA) method cannot be applied directly. This problem is known as the "small sample size" problem. In this paper, we propose a new face recognition method called the Discriminative Common Vector method based on a variation of Fisher's Linear Discriminant Analysis for the small sample size case. Two different algorithms are given to extract the discriminative common vectors representing each person in the training set of the face database. One algorithm uses the within-class scatter matrix of the samples in the training set while the other uses the subspace methods and the Gram-Schmidt orthogonalization procedure to obtain the discriminative common vectors. Then, the discriminative common vectors are used for classification of new faces. The proposed method yields an optimal solution for maximizing the modified Fisher's Linear Discriminant criterion given in the paper. Our test results show that the Discriminative Common Vector method is superior to other methods in terms of recognition accuracy, efficiency, and numerical stability.