DMCA
W.: Youdo, i-learn: Discovering task relevant objects and their modes of interaction from multi-user egocentric video
| Venue: | In: British Machine Vision Conference (BMVC) (2014 |
| Citations: | 1 - 1 self |
BibTeX
@INPROCEEDINGS{Damen_w.:youdo,,
author = {Dima Damen and Teesid Leelasawassuk and Osian Haines and Andrew Calway and Walterio Mayol-cuevas},
title = {W.: Youdo, i-learn: Discovering task relevant objects and their modes of interaction from multi-user egocentric video},
booktitle = {In: British Machine Vision Conference (BMVC) (2014},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We present a fully unsupervised approach for the discovery of i) task relevant objects and ii) how these objects have been used. Given egocentric video from multiple opera-tors, the approach can discover objects with which the users interact, both static objects such as a coffee machine as well as movable ones such as a cup. Importantly, the com-mon modes of interaction for discovered objects are also found. We investigate using appearance, position, motion and attention, and present results using each and a combi-nation of relevant features. Results show that the method is capable of discovering 95% of task relevant objects on a variety of daily tasks such as initialising a printer, preparing a coffee and setting up a gym machine. In addition, the approach enables the automatic generation of guidance video on how these objects have been used before. 1
