Learning to Detect Carried Objects with Minimal Supervision
Learning to Detect Carried Objects with Minimal Supervision
Files
Publication or External Link
Date
2012-12-21
Authors
Dondera, Radu
Morariu, Vlad I.
Davis, Larry S.
Advisor
Citation
DRUM DOI
Abstract
We propose a learning-based method for detecting carried objects that
generates candidate image regions from protrusion, color contrast and
occlusion boundary cues, and uses a classifier to filter out the regions
unlikely to be carried objects. The method achieves higher accuracy than
state of the art, which can only detect protrusions from the human
shape, and the discriminative model it builds for the silhouette
context-based region features generalizes well. To reduce annotation
effort, we investigate training the model in a Multiple Instance
Learning framework where the only available supervision is "walk" and
"carry" labels associated with intervals of human tracks, i.e., the
spatial extent of carried objects is not annotated. We present an
extension to the miSVM algorithm that uses knowledge of the fraction of
positive instances in positive bags and that scales to training sets of
hundreds of thousands of instances.