Vadim Kantorov

PhD student in computer vision

E-mail vadimkantorov@gmail.com
Phone +33603292769
Skype vadimkantorov
CV  |  GitHub  |  LinkedIn  |  Twitter  |  Facebook
photo of Vadim Kantorov

I'm a third year PhD student in computer vision with WILLOW group at INRIA / Ecole normale supérieure supervised by Ivan Laptev and Minsu Cho. I hold a masters degree in applied mathematics, computer vision and machine learning (MVA at ENS de Cachan).

My research interests these days focus on designing end-to-end deep models for solving weakly superised problems. Previously I've worked on highly efficient action recognition. Back in college I enjoyed ACM ICPC programming competitions.

Outside vision and learning, I am interested in distributed systems and street art.

Publications

ContextLocNet
Context-aware deep network models for weakly supervised localization
Vadim Kantorov, Maxime Oquab, Minsu Cho, Ivan Laptev; ECCV 2016
[paper] [poster] [bibtex] [code] [webpage]
@inproceedings{kantorov2016,
  title = {ContextLocNet: Context-aware Deep Network Models for Weakly Supervised Localization},
  author = {Kantorov, V., Oquab, M., Cho M. and Laptev, I.},
  booktitle = {Proc. European Conference on Computer Vision (ECCV), 2016},
  year = {2016}
}
fastvideofeat
Efficient feature extraction, aggregation and classification for action recognition
Vadim Kantorov, Ivan Laptev; CVPR 2014
[paper] [poster] [bibtex] [code] [webpage]
@inproceedings{kantorov2014,
  author = {Kantorov, V. and Laptev, I.},
  title = {Efficient feature extraction, encoding and classification for action recognition},
  booktitle = {Proc. Computer Vision and Pattern Recognition (CVPR), IEEE, 2014},
  year = {2014}
}
Master thesis
Evaluating MPEG compressed domain features for action recognition
Master thesis. Advisor: Ivan Laptev
[pdf]
Bachelor thesis
Automatically assessing content quality in community question answering sites
Bachelor thesis, presented at TbiLLC 2011. Advisor: Pavel Braslavski
[pdf]

Other projects

wigwam - a humane dependency fetcher and builder for scientific projects

mpegflow - a tool for extracting and dumping motion vectors from video files

caffemodel2json - a tool for dumping Caffe models to JSON for inspection

wemosetup - a Python tool for setting up Belkin WeMo lights devices without Android or iOS

vosges - a tool for defining and running Sun Grid Engine (SGE) experiments in Python