Post-doctoral position in Computer Vision - Video Analysis
Linköping University, Sweden

STARTING DATE : From September 2008 ideally

DURATION : 12 months (optional prolongation to 24 months)

SALARY (gross income) : about 2700 Euros /month. Taxation about 32%, health care included

POSITION LOCATION : Computer Vision Laboratory, Department of Electrical Engineering, Linköping University, SE-58183 Linköping, http://www.cvl.isy.liu.se

SCIENTIFIC CONTEXT : This post-doctoral position is related to the CENIIT project CAIRIS (Contents Associative Indexing and Retrieval of Image Sequences, http://www.cvl.isy.liu.se/Research/Object/CAIRIS/). The CAIRIS project is a unilateral small-scale project on extraction of generic features from image sequences for visual annotation with the purpose of later associating the features to content-based user queries.

MISSION : Assuming that the semantics of the user request to a video database is unknown a the time of indexing, no classical static indexing technique can be employed. Instead, more generic descriptions of the contents need to be generated, that can later be associated to particular semantics. The generic dimage sequence escriptors must be sufficiently rich to allow discrimination of different contents and at the same time, they must be sufficiently sparse and structured, such that fast, tree-based search in the database is feasible.

ENVIRONMENT : The Computer Vision Laboratory (CVL) has a long history in the area of Signal Processing for Computer Vision, Cognitive Vision, and Robotics. Various approaches for feature extraction, motion estimation, stereo, and tracking have been developed, e.g. double-angle representation of orientation and structure tensor. Novel information representation and learning techniques have been investigated, e.g.  channel representation and channel coded feature maps. Robust and flexible visual-servoing approaches to robot control have been proposed. CVL has been coordinator of the recently successfully completed EU FP6 project COSPAL (http://www.cospal.org) and currently coordinates the EU FP7 project DIPLECS (http://www.diplecs.eu). The relatively small team of researchers at CVL has a tradition of tight cooperation, with a particularly open climate for discussions, mutual support, and exchange of code and expertise.

REQUIRED QUALIFICATIONS : Advanced theoretical background in signal and image sequence processing and analysis, good experience of programming (C++, Matlab, and preferably GPGPU), preferably knowledge about video compression formats.

ACADEMIC BACKGROUND: 5 years higher education (university's degree or engineering equivalence). A Ph.D. in the field of computer vision, preferably image sequence processing or analysis.

CONTACTS FOR ADDITIONAL INFORMATION : Applicants should send a CV including a list of publications, a statement of research interests, and names and contact information of 3 referees electronically to Michael Felsberg, mfe@isy.liu.se before July 7, 2008.

Associate Prof. Michael Felsberg       Tel:   +46 13 282460
Computer Vision Laboratory             Fax:   +46 13 138526
Linköping University                   email: mfe@isy.liu.se
SE-581 83 Linköping, Sweden            http://www.cvl.isy.liu.se/~mfe/