Scientific Competitions

Organization and participation

We organized or participated the following scientific competitions in recent years:

2012 OrganisationICPR-2012 HARL (Human activity recognition and localization)
2016 Participation: results pending ICDAR-2015 Andar-TL Competition on Text Line Detection in Historical Documents
2014 Participation: 1st ranked (1/17) ECCV 2014 Chalearn looking at people: gesture recognition (read more)
2013 Participation: ranked 6/20 Chalearn 2013 gesture recognition
2009 Participation: ranked 5/43 ICDAR 2009 DIBCO (Document image binarization)
2003 Participation ICDAR 2003 Robust reading
2002 Participation TrecVid 2002
2015 Contribution of the metric ICDAR 2015 Robust reading
2013 Contribution of the metric ICDAR 2013 Robust reading
2011 Contribution of the metric ICDAR 2011 Robust reading
2007 Contribution of the metric ImageEval 2007
2003 Contribution of the metric ICDAR 2003 Robust reading

Evaluation of object detection algorithms

We have proposed a new performance graphs for the evalaluation of object detection algorithms, which illustrate performance intuitively by graphs which present object level precision and recall depending on constraints on detection quality. In order to compare different detection algorithms, a representative single performance value is computed from the graphs. The evaluation method can be applied to different types of object detection algorithms. It has been tested on different text detection algorithms, among which are the participants of the ICDAR 2003 text detection competition.

This method has been used to evaluate the participants in several scientific competitions:

  • ICDAR 2015 Robust reading competition [WWW]
  • ICDAR 2013 Robust reading competition [WWW]
  • ICDAR 2011 Robust reading competition [WWW]
  • Image Eval 2007 text detection competition [REF]
  • ICDAR 2003 Robust Reading Competition [REF]
  • Christian Wolf and Jean-Michel Jolion. Object count/Area Graphs for the Evaluation of Object Detection and Segmentation Algorithms, In International Journal on Document Analysis and Recognition , 8(4):280-296, 2006.

Evaluation of activity recognition and localization algorithms

We have presented a new performance metric for the evaluation of activity recognition and localization algorithms. The method has been used to evaluate the participants of the ICPR HARL 2012 competition.



Christian Wolf
INSA-Lyon / Université de Lyon