Using body-anchored priors for identifying actions in single images.

Leonid Karlinsky Michael Dinerstein Shimon Ullman

Published in: NIPS (2010)

Keyphrases

image data
ground truth
image database
input image
rigid body
image collections
test images
image classification
image features
object recognition
image matching
three dimensional
image analysis
image retrieval
image pixels
image annotation
edge detection
image registration
multiple images
image content
illumination conditions
original images
image set
prior information
lighting conditions
region of interest
keypoints
segmentation method
single image
natural images