[Publications] [Awards] [Lectures/Talks] [Theses] [Work Experience] [Education]


Berk Gökberk
University of Twente
Signals and Systems Group,
Department of Electrical Engineering Mathematics and Computer Science
Postal address:
Signals and Systems Group
Electrical Engineering
University of Twente
P.O.Box 217
7500 AE Enschede
The Netherlands
 		Visiting address:
Signals and Systems Group
University of Twente
Hogekamp Building (building no. 45)
de Veldmaat 10
7522 NB Enschede
The Netherlands

Publications
Invited Talks & Lectures
Events Organized
Awards
Summer Schools/Workshops

Biosecure First Residential Workshop,
Paris, France, August 2005

Biometrics Workshop,
Challenges arising from Theory to Practice (BCTP), Cambridge, UK, August 22th, 2004. [link]

Advanced Studies in Biometrics,
Summer School on Biometrics, Alghero, Italy, June 2-6, 2003.

Theses
Ph.D. Thesis
Title: Three-Dimensional Face Recognition [pdf]
Advisors: Prof. Lale Akarun, Prof. Bülent Sankur, Prof. Fikret Gürgen, Prof. Ethem Alpaydın
Abstract: In this thesis, we attack the problem of identifying humans from their three dimensional facial characteristics. For this purpose, a complete 3D face recognition system is developed. We divide the whole system into sub-processes. These sub-processes can be categorized as follows: 1) registration, 2) representation of faces, 3) extraction of discriminative features, and 4) fusion of matchers. For each module, we evaluate the state-of-the art methods, and also propose novel ones. For the registration task, we propose to use a generic face model which speeds up the correspondence establishment process. We compare the benefits of rigid and non-rigid registration schemes using a generic face model. In terms of face representation schemes, we implement a diverse range of approaches such as point clouds, curvature-based descriptors, and range images. In relation to these, various feature extraction methods are used to determine the discriminative facial features. We also propose to use local region-based representation schemes which may be advantageous in terms of both dimensionality reduction and for determining invariant regions under several facial variations. Finally, with the realization of diverse 3D face experts, we perform an in-depth analysis of decision-level fusion algorithms. In addition to the evaluation of baseline fusion methods, we propose to use two novel fusion schemes where the first one employs a confidence-aided combination approach, and the second one implements a two-level serial integration method. Recognition simulations performed on the 3DRMA and the FRGC databases show that: 1) generic face template-based rigid registration of faces is better than the non-rigid variant, 2) principal curvature directions and surface normals have better discriminative power, 3) representing faces using local patch descriptors can both reduce the feature dimensionality and improve the identification rate, and 4) confidence-assisted fusion rules and serial two-stage fusion schemes have a potential to improve the accuracy when compared to other decision-level fusion rules.


M.S. Thesis
Title: Feature-based Pose Invariant Face Recognition [pdf]
Advisors:Prof. Lale Akarun, Prof. Ethem Alpaydın
Abstract: One of the major difficulties in face recognition systems is the in-depth pose variation problem. Most face recognition approaches assume that the pose of the face is known. In this work we compared two different face representation methods for pose invariant face recognition. The first one is the conventional Eigenface coding of face images. The second one is based on the Gabor wavelet based filtering of face images. Gabor wavelets are chosen because of their sensitivity to both spatial frequency and orientation, like the early stages of human visual processing system. Subspace coding of Gabor wavelet-based face representation was also implemented using Principal Component Analysis without a significant loss in the recognition performance of the system. Both parametric and view-based pose invariant face recognition algorithms were used for each representation. Our results show that Gabor wavelet based filtering of images improves the overall performances of both parametric and view-based approaches where these two classification methods have almost similar recognition performances. Effectiveness of Gabor wavelet-based representation was also shown for frontal face databases.

Work Experience

1998 - 1999
Oracle Turkey, Techinal Analyst

Education
2001 - 2006,  Ph.D., Boğaziçi University, Computer Engineering Department
1999 - 2001,  M.S., Boğaziçi University, Computer Engineering Department
1994 - 1999,  B.S., Boğaziçi University, Computer Engineering Department
Past Projects

Biosecure Network of Excellence (NoE) European 6th Framework Project,
Development of 2D/3D face recognition/authentication reference systems
Biosecure Joint Project: 3D-assisted 2D and 3D face registration and matching,
Project Partners: Boğaziçi University, University of Surrey, University of Sassari
3D Object Recognition,
The Scientific and Technological Research Council of Turkey (TUBITAK), Project Code: 104E080
Perceptual Human Computer Interaction,
T.R. Prime Ministry State Planning Organization (DPT), Project Code: 03K120250

Other Activities
Last update: November, 2008
Best viewed with 44Khz, 256kps Starless and Bible Black -King Crimson