Prof. Marios Savvides (left),
Prof. Vijayakumar Bhagavatula (middle)
receive Oustanding CIT Research Award from Dean Pradeep Khosla (right)
Oustanding CIT Research Award for Winning Frequency Based Face and Iris Recognition Algorithms
Prof. Marios Savvides and Prof. Vijayakumar Bhagavatula receive the CMU Carnegie Institute of Technology (CIT) Outstanding Research Award for their award winning Frequency based Face Recognition algorithms and Iris recognition algorithms that landed them 1st place in NIST's Face Recognition Grand Challenge (FRGC) 2006 Open Challenge, a world-wide competition involving Academics and Industry Vendors across the Globe.
Figure above from NIST report shows we outperformed commercial vendors (median green line) almost double the face recognition performance on the hardest experiment #4.
NIST Iris Challenge Evaluation 2006 Results
We additionally were asked to participate and submit our iris algorithm to NIST's Iris Challenge Evaluation (ICE), on the open challenge, we landed a 3-way top performance with the 2 top performing vendors (French based SAGEM Morpho) and Korean based Iritech companies. On the left eye experiment denoted by blue bar below, Iritech was #1, we(CMU) were #2 and SAGEM was #3. On the right eye experiment denoted by the red bar below, SAGEM was #1, we CMU were #2, and Iritech was #3, hence we were on a 3-way tie with SAGEM and Iritech.
We were the ONLY group that partipated in BOTH competitions, SIMULTANEOUSLY and landed TOP performing algorithms!!
NIST Face Recognition Grand Challenge 2006 Results
PITTSBURGH—Carnegie Mellon University's Marios Savvides is one of four researchers selected to be part of a new Center for Academic Studies In Identity Sciences (CASIS), a Center of Academic Excellence (CAE), under the Office of the Director of National Intelligence (ODNI). This is the ODNI's first CAE created as a pilot with an emphasis on science and technology in the area of Biometrics. Our work in Biometrics has really taken off and recognized within the community, we are proud of the technological advances we have made!
1st place in hand on wheel detection and hand classification (May. 2018)
The Vision for Intelligent Vehicles and Applications (VIVA) Challenge
Our proposed MS-FRCN approach consistently achieves the state-of-the-art hand detection results, i.e. Average Precision (AP) / Average Recall (AR) of 95.1% / 94.5% at level 1 and 86.0% / 83.4% at level 2, on the VIVA challenge
The proposed method achieves the state-of-the-art results for left/right hand and driver/passenger classification tasks on the VIVA database with a significant improvement on AP/AR of 7% and 13%
for both classification tasks, respectively.