Matching Items (2)
Description
In this thesis we consider the problem of facial expression recognition (FER) from video sequences. Our method is based on subspace representations and Grassmann manifold based learning. We use Local Binary Pattern (LBP) at the frame level for representing the facial features. Next we develop a model to represent the video sequence in a lower dimensional expression subspace and also as a linear dynamical system using Autoregressive Moving Average (ARMA) model. As these subspaces lie on Grassmann space, we use Grassmann manifold based learning techniques such as kernel Fisher Discriminant Analysis with Grassmann kernels for classification. We consider six expressions namely, Angry (AN), Disgust (Di), Fear (Fe), Happy (Ha), Sadness (Sa) and Surprise (Su) for classification. We perform experiments on extended Cohn-Kanade (CK+) facial expression database to evaluate the expression recognition performance. Our method demonstrates good expression recognition performance outperforming other state of the art FER algorithms. We achieve an average recognition accuracy of 97.41% using a method based on expression subspace, kernel-FDA and Support Vector Machines (SVM) classifier. By using a simpler classifier, 1-Nearest Neighbor (1-NN) along with kernel-FDA, we achieve a recognition accuracy of 97.09%. We find that to process a group of 19 frames in a video sequence, LBP feature extraction requires majority of computation time (97 %) which is about 1.662 seconds on the Intel Core i3, dual core platform. However when only 3 frames (onset, middle and peak) of a video sequence are used, the computational complexity is reduced by about 83.75 % to 260 milliseconds at the expense of drop in the recognition accuracy to 92.88 %.
ContributorsYellamraju, Anirudh (Author) / Chakrabarti, Chaitali (Thesis advisor) / Turaga, Pavan (Thesis advisor) / Karam, Lina (Committee member) / Arizona State University (Publisher)
Created2014
Description
Broaden and build theory (BBT; Fredrickson, 1998; 2001) postulates that positive emotions expand the scope of one's attention and thought-action repertoires (Fredrickson & Branigan, 2005). Within the boundaries of BBT, the undoing hypothesis (Fredrickson, 1998, Fredrickson & Levenson, 1998) argues that positive emotions themselves do not bring forth specific action tendencies or urges; therefore, they do not consequently require an increase in cardiovascular activity to carry out the urge. On the other hand, positive emotions have evolved to subdue the cardiovascular response previously initiated by negative emotions. This dissertation proposes that the real power of positive emotions might be to undo not the effects of negative emotions themselves, however, but simply reduce the arousal itself. This dissertation used minor physiological arousal (e.g., a step-stool task) to simulate the cardiovascular effects of the stress manipulations used in previous tests of the undoing hypothesis by Fredrickson and colleagues. This dissertation asks if positive emotions undo the cardiovascular reactivity of an emotionally neutral stimulus. Positive emotions were induced through one film clip (i.e., a happy film clip) and was compared to a neutral film clip (no emotion elicited). An experimental design measured the effects of arousal induction and film clip on participants' cardiovascular activity. Results indicated that positive emotions had the same effect as no emotions on participants' cardiovascular activity. Implications for theory and research are provided, as well as an assessment of the study's strengths and limitations. Finally, several directions for future research are offered.
ContributorsDeiss, Douglas M (Author) / Floyd, Kory (Thesis advisor) / Mongeau, Paul (Committee member) / Thompson, Marilyn (Committee member) / Arizona State University (Publisher)
Created2012