Oxford U: PDRA Position in Computer Vision

Oxford U: PDRA Position in Computer Vision

Posted by Rebecca Martin on Tue, 29/07/2014 - 14:35


Postdoctoral research assistant in detailed and deep image understanding Starting preferably before August 31st, 2014 or as soon as possible thereafter

We are seeking a Postdoctoral research assistant to join the Visual Geometry Group (VGG) in the Oxford Engineering Department. The post is funded by the UK Engineering and Physical Sciences Research Council (EPSRC), and is fixed-term up to 12 months.

The candidate will work on a cutting-edge research project at the boundary of computer vision and deep learning. In the past few years, deep learning has dramatically outperformed conventional computer vision methods in key visual tasks such as object recognition, detection, and segmentation. However, empirical evidence suggests that our understanding of these very promising models is still limited. The primary aim of this project is to characterise the information captured by a deep architectures learned on very large datasets, focusing in particular on discovering visual semantics as may be captured by the model and how the latter is represented and acquired during learning.

The candidate will carry theoretical as well as empirical work on deep learning applied to computer vision tasks and may co-supervise DPhil students working on related problems.

The candidate should posses (or be near completion of) a relevant PhD/DPhil in computer vision or machine learning, together with a solid mathematical and computational background in learning, optimization, and vision. Previous experience on the practical aspects of the large-scale numerical methods at the core of this research is welcome. A solid track record of publications in relevant international peer-reviewed conferences and journals (e.g. ICCV, CVPR, ECCV, NIPS, ICML, ICLR, PAMI, IJCV) is essential.

For informal enquiries please contact vedaldi(at)robots.ox.ac.uk.

On applying, you will be required to upload a cover letter, a brief statement of research interest (describing how past experience and future plans fit with the advertised position), CV, and details of two referees.

About the Oxford Visual Geometry Group. The research will be conducted in the VGG within the Robotic Research Group at the University of Oxford, which has an international reputation for large scale recognition of images and videos, and for applications of machine learning to computer vision. The candidate will have the opportunity of interacting with leading computer vision scientists such as Andrew Zisserman, David Murray and Philip Torr, including participating in joint reading groups and other joint research activities.

Contact information
Contact Person: Prof Andrea Vedaldi
Contant Email: vedaldi(at)robots.ox.ac.uk