Advanced Multimedia Processing Lab -- Projects -- Personal Authentication Based on SMMS

            About AMP Lab        Projects        Downloads        Publications        People        Links

Project - Personal Authentication Based on SMMS

 

Top of this page

                                  

Motivation and Goal

As the biometric applications become more popular, the biometric classification algorithms will be applied to hand-held devices. Due to the size and power constraints of these devices, the computational complexity of the verification procedure will be the main issue.  

In this project, we want to study how we can build a new classifier for authentication which has the simple computational requirement on verification with the good performance. We propose a new classification algorithm based on the concept of Symmetric Maximized Minimal distance in Subspace (SMMS). 

Top of this page

 

Symmetric Maximized Minimal distance in Subspace (SMMS)

We introduce a new classification algorithm based on the concept of Symmetric Maximized Minimal distance in Subspace (SMMS). Given the training data of authentic samples and imposter samples in the feature space, SMMS tries to identify a subspace in which all the authentic samples are clustered together and all the imposter samples are far away from the authentic samples.  The optimality of the subspace is determined by maximizing the minimal distance between the authentic samples and the imposter samples in the subspace.  

We present a procedure to achieve such optimality and to identify the subspace and the decision boundary.  Once the subspace is trained, the verification procedure is simple since we only need to project the test sample onto the subspace and compare it against the decision boundary.  

Using face authentication as an example, we show that the proposed algorithm outperforms the linear classifier based on Support-Vector Machines (SVM). The proposed algorithm also applies to multimodal feature spaces.  The features can come from any modalities, such as face images, voices, fingerprints, etc.

Top of this page

 

 

Top of this page

 

Any suggestions or comments are welcome. Please send them to Wende Zhang. 

Top of this page


This website is maintained by Wende Zhang
Copyright 200
1 Advanced Multimedia Processing Lab. All rights reserved.
Revised: March 08, 2004 .