Detail výsledku
Face Representation and Tracking Using Gabor Wavelet Networks
This work presents a one of the approaches to human gesturerecognition. A method where a discrete face template is represented bylinear combination of the 2D odd-Gabor wavelet functions (Gabor WaveletNetwork) is proposed. Using wavelet networks, an effective facerepresentation and tracking method is achieved that is robust tochanges and deformations of the face image. Possibilities ofcooperation between a GWN and others methods for robust facialexpressions recognition are also discussed, for example: PrincipalComponent Analysis (PCA) and parameterized models of optical flow.
Image, Object representation, Feature-based, Template-based, 2D Gaborwavelet functions, Gabor Wavelet Networks, Facial features, Tracking,Facial expressions, Parameterized models of optical flow, PrincipalComponent Analysis
@inproceedings{BUT10927,
author="Michal {Španěl} and Pavel {Zemčík}",
title="Face Representation and Tracking Using Gabor Wavelet Networks",
booktitle="Proceedings of the Central European Seminar on Computer Graphics",
year="2003",
pages="163--169",
address="Budmerice",
url="https://www.fit.vut.cz/research/publication/7366/"
}