Date of Award
Electrical and Computer Engineering
Wu, Jonathan (Electrical and Computer Engineering)
Engineering, Electronics and Electrical.
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Facial expression recognition from images or videos attracts interest of research community owing to its applications in human-computer interaction and intelligent transportation systems. The expressions cause non-rigid motions of the face-muscles thereby changing the orientations of facial curves. Wavelets and Gabor wavelets have been used effectively for recognition of these oriented features. Although wavelets are the most popular multiresolution method, they have limited orientation-selectivity/directionality. Gabor wavelets are highly directional but they are not multiresolution methods in the true sense of the term. Proposed work is an effort to apply directional multiresolution representations like curvelets and contourlets to explore the multiresolution space in multiple ways for extracting effective facial features. Extensive comparisons between different multiresolution transforms and state of the art methods are provided to demonstrate the promise of the work. The problem of drowsiness detection, a special case of expression recognition, is also addressed using a proposed feature extraction method.
Saha, Ashirbani, "Facial Expression Recognition Using Multiresolution Analysis" (2010). Electronic Theses and Dissertations. 137.