Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/28552
Citations | ||
Scopus | Web of Science® | Altmetric |
---|---|---|
?
|
?
|
Type: | Conference paper |
Title: | On accuracy/robustness/complexity trade-offs in face verification |
Author: | Sanderson, Conrad Cardinaux, Fabien Bengio, Samy |
Citation: | Proceedings of the Third International Conference on Information Technology and Applications / IEEE:pp.638-643 |
Publisher: | IEEE Computer Society |
Issue Date: | 2005 |
ISBN: | 0769523161 |
Conference Name: | International Conference on Information Technology and Applications (3rd : 2005 : Sydney, Australia) ICITA’05 |
School/Discipline: | School of Electrical and Electronic Engineering |
Statement of Responsibility: | Conrad Sanderson, Fabien Cardinaux, Samy Bengio |
Abstract: | In much of the literature devoted to face recognition, experiments are performed with controlled images (e.g. manual face localization, controlled lighting, background and pose). However, a practical recognition system has to be robust to more challenging conditions. In this paper we first evaluate, on the relatively difficult BANCA database, the discrimination accuracy, robustness and complexity of Gaussian Mixture Model (GMM), 1D- and pseudo-2D Hidden Markov Model (HMM) based systems, using both manual and automatic face localization. We also propose to extend the GMM approach through the use of local features with embedded positional information, increasing accuracy without sacrificing its low complexity. Experiments show that good accuracy on manually located faces is not necessarily indicative of good accuracy on automatically located faces (which are imperfectly located). The deciding factor is shown to be the degree of constraints placed on spatial relations between face parts. Methods which utilize rigid constraints have poor robustness compared to methods which have relaxed constraints. Furthermore, we show that while the pseudo-2D HMM approach has the best overall accuracy, classification time on current hardware makes it impractical. The best trade-off in terms of complexity, robustness and discrimination accuracy is achieved by the extended GMM approach. |
Description: | Copyright © 2005 IEEE |
DOI: | 10.1109/ICITA.2005.192 |
Appears in Collections: | Electrical and Electronic Engineering publications |
Files in This Item:
File | Size | Format | |
---|---|---|---|
hdl_28552.pdf | 543.77 kB | Publisher's PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.