Action Localization by Tubelets from Motion M. Jain, J. C. van Gemert, H. Jégou, P. Bouthemy, C. G. M. Snoek In IEEE Conference on Computer Vision and Pattern Recognition 2014. [bibtex] [pdf] [url]

This paper considers the problem of action localization, where the objective is to determine when and where certain actions appear. We introduce a sampling strategy to produce 2D+t sequences of bounding boxes, called tubelets. Compared to state-of-the-art alternatives, this drastically reduces the number of hypotheses that are likely to include the action of interest. Our method is inspired by a recent technique introduced in the context of image localization. Beyond considering this technique for the first time for videos, we revisit this strategy for 2D+t sequences obtained from super-voxels. Our sampling strategy advantageously exploits a criterion that reflects how action related motion deviates from background motion. We demonstrate the interest of our approach by extensive experiments on two public datasets: UCF Sports and MSR-II. Our approach significantly outperforms the state-of-theart on both datasets, while restricting the search of actions to a fraction of possible bounding box sequences.

@InProceedings{JainCVPR2014,
  author       = "Jain, M. and van Gemert, J. C. and J\'egou, H. and Bouthemy, P.
                  and Snoek, C. G. M.",
  title        = "Action Localization by Tubelets from Motion",
  booktitle    = "IEEE Conference on Computer Vision and Pattern Recognition",
  year         = "2014",
  url          = "https://ivi.fnwi.uva.nl/isis/publications/2014/JainCVPR2014",
  pdf          = "https://ivi.fnwi.uva.nl/isis/publications/2014/JainCVPR2014/JainCVPR2014.pdf",
  has_image    = 1
}