Stavros Tsogkas

MaRS Discovery District
101 College St., M5G 1L7, Toronto, Ontario, Canada

/ / /

I am a Research Scientist at the Samsung Artificial Intelligence Center (SAIC) in Toronto and an affiliate member at the Vector Institute for Artificial Intelligence. Before that I was a Postdoctoral Researcher at the University of Toronto working with Sven Dickinson. Before joining UofT I spent five years as a PhD student and Research Engineer at CentraleSupélec, where I was advised by Iasonas Kokkinos. My alma mater is the National Technical University of Athens.

Research Interests

My research interests are in the broad areas of computer vision and machine learning, with a focus on deep learning. I am particularly interested in the use of mid-level representations to bridge the gap between bottom-up and top-down processing and solve problems such as object detection, segmentation and grouping. I have devoted a large part of my research on recovering such representations, medial axes and object parts in particular, in natural and medical images.



DeepFlux for Skeletons in the Wild
Y. Wang, Y. Xu, S. Tsogkas, X. Bei, S. Dickinson, K. Siddiqi
Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR 2019).
code / bibtex

2017 ICCV Challenge: Detecting Symmetry in the Wild (editorial)
C. Funk*, S. Lee*, M. Oswald*, S. Tsogkas*, W. Shen, A. Cohen, S. Dickinson, Y. Liu
Detecting Symmetry in the Wild Workshop (In conjunction with ICCV 2017).
*Authors contributed equally

AMAT: Medial Axis Transform for Natural Images
S. Tsogkas, S. Dickinson
International Conference on Computer Vision (ICCV 2017).
code / bibtex / video / slides

Prior-based Coregistration and Cosegmentation
M. Shakeri*, E. Ferrante*, S. Tsogkas, S. Lippe, S. Kadoury, I. Kokkinos, N. Paragios
International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2016).
*Authors contributed equally

Subcortical Brain Structure Segmentation Using FCNNs
S. Tsogkas*, M. Shakeri*, E. Ferrante, S. Lippe, S. Kadoury, N. Paragios, I. Kokkinos
International Symposium on Biomedical Imaging (ISBI 2016) - oral.
*Authors contributed equally
code / slides (pdf) / slides (odp) / bibtex

Accurate Human-Limb Segmentation in RGB-D images for Intelligent Mobility Assistance Robots
S. Chandra, S. Tsogkas, I. Kokkinos
Third Workshop on Assistive Computer Vision and Robotics (In conjunction with ICCV 2015).
poster / bibtex

Deep Learning for Semantic Part Segmentation with High-Level Guidance
S. Tsogkas, I. Kokkinos, G. Papandreou, A. Vedaldi
arXiv report.

Deformable Part Models with CNN Features
P. A. Savalle, S. Tsogkas, G. Papandreou, I. Kokkinos
European Conf. on Computer Vision, Parts and Attributes Workshop (ECCVW 2014).
poster / bibtex

Segmentation-aware Deformable Part Models
E. Trulls, S. Tsogkas, I. Kokkinos, A. Sanfeliu, F. Moreno
Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR 2014).
poster / bibtex

Understanding Objects in Detail with Fine-grained Attributes
A. Vedaldi, S. Mahendran , S. Tsogkas, S. Maji, R.B. Girshick, J. Kannala, E. Rahtu, I. Kokkinos,
M. B. Blaschko, D. Weiss, B. Taskar, K. Simonyan, N. Saphra, S. Mohamed

Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR 2014).
code / data / poster / bibtex

Learning-Based Symmetry Detection in Natural Images
S. Tsogkas , I. Kokkinos
Proc. European Conf. on Computer Vision (ECCV 2012).
code / data / poster / bibtex

Ph.D thesis

Mid-level Representations for Modeling Objects
PhD Thesis, Advisor: I. Kokkinos
Center for Visual Computing, CentraleSupelec.