DMCA
Watching Unlabeled Video Helps Learn New Human Actions from Very Few Labeled Snapshots
| Citations: | 8 - 1 self |
BibTeX
@MISC{Chen_watchingunlabeled,
author = {Chao-yeh Chen and Kristen Grauman},
title = {Watching Unlabeled Video Helps Learn New Human Actions from Very Few Labeled Snapshots},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We propose an approach to learn action categories from static images that leverages prior observations of generic human motion to augment its training process. Using unla-beled video containing various human activities, the system first learns how body pose tends to change locally in time. Then, given a small number of labeled static images, it uses that model to extrapolate beyond the given exemplars and generate “synthetic ” training examples—poses that could link the observed images and/or immediately precede or follow them in time. In this way, we expand the training set without requiring additional manually labeled exam-ples. We explore both example-based and manifold-based methods to implement our idea. Applying our approach to recognize actions in both images and video, we show it en-hances a state-of-the-art technique when very few labeled training examples are available. 1.
Keyphrases
labeled snapshot manifold-based method prior observation synthetic training example body pose tends unla-beled video small number training process generic human motion action category static image observed image state-of-the-art technique various human activity training example labeled static image
