Curriculum Vita
Curriculum Vita (last update 2/12/2013). Google Scholar Profile
One Paragraph Bio. (last updated 2/10/2013)
Professor Beveridge works on Computer Vision, emphasizing problems relating to object recognition. Current interests include evaluation methodology and the evaluation of human face recognition algorithms in particular. Open source tools for benchmarking face recognition algorithm performance are an important component of his face recognition work. Two recent baseline algorithms have been released, and the older CSU Face Identification Evaluation System has been downloaded well over 20,000 times since its introduction in 2001. The face recognition work has also resulted in the largest covariate studies to date exploring what factors make a person's face harder or easier to recognize using standard algorithms. In 2009, an open source live video face recogntion system, FaceL, was released to the web that incorporates new innovations in the development of correlation filters. Professor Beveridge's other interests include high dimensional data analysis, optimal matching of geometric features, genetic algorithms and the use reconfigurable embedded hardware.
Short Bio. (last updated 3/10/2011)
J. Ross Beveridge received his B.S. degree in Applied Mechanics and Engineering Science from the University of California at San Diego in 1980 and his M.S. and Ph.D. degrees in Computer Science from the University of Massachusetts in 1987 and 1993 respectively. He has been in the Computer Science Department at Colorado State University since 1993, where he was an Assistant Professor from 1993 to 2000, an Associate Professor from 2000 to 2010, and where he is currently a Professor.
Dr. Beveridge is a member of the IEEE Computer Society as well as the ACM. He has served as an Associate Editor for the IEEE Transactions on Pattern Recognition and Machine Intelligence (PAMI), Pattern Recognition and Image and Vision Computing. He was Program Co-Chair for the 1999 IEEE Conference on Computer Vision and Pattern Recognition and frequently serves on numerous workshop and conference Program Committees. He has received Outstanding Reviewer awards in 2007 and 2008 from the organizers of the IEEE Conference on Computer Vision and Pattern Recognition. He is the author of over 100 publications in computer vision and related fields.
Recent research accomplishments include developing reference implementations of standard human face recognition algorithms and associated experimental protocols to characterize uncertainty in common performance measures such as recognition rate. These algorithms and protocols are included in the CSU Face Identification Evaluation System, an open source tool that has been downloaded over 20,000 times since its introduction in 2001. Dr. Beveridge, working with colleagues Dr. Geof Givens in Statistics and Bruce Draper in Computer Science, has also executed the largest covariate studies to date examining what combinations of factors associated with a person make that person's face harder or easier to recognize using standard face recognition algorithms. He and his colleagues at CSU are part of the support team for the NIST run Multiple Biometrics Grand Challenge. His PhD student, David Bolme, has released an open source face recognition system, FaceL, that anyone can download and use to play with live face recognition over video.
From 1993 through 1998, Dr. Beveridge developed novel 3D object recognition algorithms that fuse sensor data through a 3D model and iteratively refine an instantiated scene model through repeated rendering and matching between images and model. In the context of Automatic Target Recognition, these algorithms were tested against a data set jointly collected by Colorado State and Lockheed Martin. This "Fort Carson Data Set" includes IR, color and range imagery available through the Colorado State web site.
Other areas of past or current research include optimal matching of geometric features, genetic algorithms and evolutionary computation, the use of reconfigurable embedded hardware for image processing, behavior learning for unmanned air vehicles and, most recently, severe weather prediction in GOES-R sounder data.
