| Department |
Electrical and Computer Engineering |
|---|---|
| Office | B209 Hamerschlag Hall |
| Telephone | (412)-268-1142 |
| Fax | (412)-268-6345 |
| marioss@andrew.cmu.edu | |
| Website | http://www.cs.cmu.edu/~msavvid/ |
| Assistant | Kelley Conley |
My research is focused on how to recognize a person based on their biometrics, e.g. their face, iris, fingerprint or palm print. These biometrics offer characteristics unique to each individual. Most of the security systems rely on passwords, swipe cards, and RF ID tags to recognize and grant access to users. However these can easily be lost or stolen, which can lead to intruders gaining access to secure physical and virtual spaces. These problems are overcome if one uses biometrics as a method to authenticate a user's identity. Biometric recognition has many challenges as the appearance of the biometrics can vary at the time of acquisition, e.g. in face recognition, a person's face can appear different due to variations in lighting, change in pose and expression. My research is focused on developing robust algorithms that can handle these type of intra-class variations with different biometric modalities. I am also researching reducing the complexity of these algorithms so that they can be implemented on small form-factor devices such as PDAs and cell-phones.
Signals and Systems
Biometric identification, Cyber-Physical Security, Authentication, Pattern recognition
PhD, 2004
Electrical and Computer Engineering
Carnegie Mellon University
MS, 2000
Robotics
Carnegie Mellon University
BEng, 1997
Microelectronics Systems Engineering
University of Manchester Institute of Science and Technology
 from July 5-29. The instructors and their topics include: Hyong Kim, "A New Direction for Next Generation Network Architecture;" Greg Ganger, "'Self-*' Storage Infrastructures;" Adrian Perrig, "Secure Sensor Networks;" Mike Reiter, "Overview of Core Security Courses;" and Marios Savvides, "The Application of Biometric Technologies." Kim directs CyLab Korea and Reiter is the technical director for CyLab. <br /> <br />The IACBP, funded by the National Science Foundation (NSF), is an intensive in-residence summer program to help build information assurance education and research capacity at home institutions, specifically minority-serving universities.</p> false asset related-people page directory/department/faculty/G/Greg_Ganger_54 false 9a41499180028154045ca9fddbf66f8a asset related-people page directory/department/faculty/K/Hyong_Kim_81 false 9a433a7680028154045ca9fd0d9155b6 asset related-people page directory/department/faculty/P/Adrian_Perrig_138 false 9a463f4980028154045ca9fdb2ad5afe asset related-people page false asset related-people page directory/department/faculty/S/Marios_Savvides_275 false 9a4b924280028154045ca9fd983586f4 asset _media/news-images/2005/07/nsf4c.gif e74c2547800281541caa270d4e384bbf image file false group text group-name Carnegie Mellon CyLab false text group-url http://www.cylab.cmu.edu/ false text group-abbr CyLab false related-groups false group text group-name Information Networking Institute false text group-url http://www.ini.cmu.edu/ false text group-abbr INI false related-groups false ){kind=link}
 held in Taiwan this spring. Li's advisor is Marios Savvides, assistant research professor of ECE. Li is a graduate student in the Language Technologies Institute.</p> <p>Their paper, titled "A Pixel-Wise, Learning-Based Approach For Occlusion Estimation of Iris Images in a Polar Domain," proposes a novel way to solve the problem of iris mask estimation which has proven to be both efficient and accurate compared to existing methods in the literature.</p> <p>In iris recognition, before researchers can employ pattern recognition, they need to perform iris segmentation and iris mask estimation. The goal of iris segmentation is to localize the eye in a given image. The goal of iris mask estimation is to estimate which regions are true iris texture and which are not. Once the true iris region is determined, researchers can then perform pattern recognition. If the iris masks are not estimated correctly, no matter how good the feature extraction and matching algorithms are, the iris recognition performance would be seriously degraded. </p> <p>In this work, the authors treat the problem of iris mask generation as a machine learning problem, which is a completely new perspective in this field. By employing a Gaussian Mixture Model learning technique on local image features, they are able to successfully create an iris mask which is accurate (compared to manually created mask) and also very efficient (compared to other existing methods). Experimental results also show the created mask substantially enhances the iris recognition performance.</p> false asset related-people page directory/department/faculty/S/Marios_Savvides_275 false 9a4b924280028154045ca9fd983586f4 asset _media/news-images/2009/06/Li_Yung-hui.jpg e75e0dee800281541caa270da3a79207 image file false asset _media/news-images/2009/06/marioss.jpg e75e12a9800281541caa270d1ac85617 image file false ){kind=link}
![Keeping an Eye on the Bad Guy]( text body <p>Assistant Research Professor Marios Savvides' iris recognition research was recently featured in a July 28 article in <a href="http://www.scientificamerican.com/article.cfm?id=hands-free-iris-biometrics"><em>Scientific American</em></a>. The government plans on having the biometric security systems that can identify individuals based on the unique patterns in their irises in the hands of the military and law enforcement by 2012.</p> <p>And across the pond, the research was featured on <a href="http://fwd.five.tv/gadget-show/videos/other/biometrics"><em>The Gadget Show</em></a>, the popular British television series that focuses on the world of technology.</p> false asset related-people page directory/department/faculty/S/Marios_Savvides_275 false 9a4b924280028154045ca9fd983586f4 asset _media/news-images/2009/08/marioss.jpg e75e9fe0800281541caa270db8f70dd3 image file false ){kind=link}
![A Look Inside New Security Research]( text body <p>"Just one look, that's all it took..." goes the old R&B classic. ECE research professor Marios Savvides is proving that true again today, although for an entirely different purpose.</p> <p>Savvides' Biometrics Research Lab is the first academic setting to host a state-of-the-art iris recognition system. Iris-on-the-Move (IOM) can acquire iris images unobtrusively at three meters as a person walks through a portal. Savvides' research has pushed this limit to a 13 meter stand-off yielding a 4X factor in acquisition improvement.</p> <p>Pittsburgh's WTAE News got <a href="http://www.thepittsburghchannel.com/news/18980382/detail.html">a look inside</a> Savvides la and walks viewers through some exciting, futuristic research.</p> false asset related-people page directory/department/faculty/S/Marios_Savvides_275 false 9a4b924280028154045ca9fd983586f4 asset _media/news-images/2009/03/marioss.jpg e75d1b54800281541caa270dddf118b8 image file false ){kind=link}