Videmus nunc per speculum in enigmate: tunc autem facie ad faciem;
nunc cognosco ex parte: tunc autem cognoscam sicut et cognitus sum.
[ I Corinthians, 13:12 ]

CBA Research is a partly professional partly personal website, a modest interface to my quest in science and life. Here you'll find synopses of my current and past research projects and related publications, my courses as well as some personal thoughts on research.

Highlights

I am one of the founders of Vispera Information Technologies. In Vispera, we design and develop visual recognition solutions for the retail industry.

I am an investor and Technical Advisor at Flavourspace, an Amsterdam-based food search and discovery startup.

I am Research Associate at Boğaziçi University EE/BUSIM and at BU/EE VAVlab research labs. I am also Adjunct Faculty at the EE Dept. in Bogazici University and part-time Lecturer at the Architectural Design Program in Istanbul Technical University -see courses.

During 2010-2014, I was with Vistek ISRA Vision as R&D Manager. Before joining Vistek ISRA Vision, I was Marie Curie postdoctoral fellow at Philips Research Europe (Video Processing and Analysis Group).

I did my PhD at Télécom ParisTech Signals-Images and at Boğaziçi University EE/BUSIM, where I developed shape description and similarity learning algorithms for content-based 3D object retrieval. For a short while after my PhD, I worked as postdoc at Bogazici University EE/VAVlab. My BS and MS degrees are both in EE and both from Boğaziçi University (2002 and 2004).

For more information, navigate through projects and publications, or download resume.

Research Interests

Content-Based Retrieval: multimedia content description, relevance feedback and models, user modeling, intelligent search paradigms.

Machine Learning: similarity learning, statistical ranking, discriminative models, active learning, data mining, collaborative filtering, semi-supervised learning, Bayesian theory and its applications.

Computer Vision: visual content analysis and description, image and shape matching, object recognition, image segmentation.