Abstract: A core challenge in action recognition from videos is obtaining sufficient training examples to train deep networks. This holds especially for action tasks from non-standard sensors such as ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback