Dense saliency-based spatiotemporal feature points for action recognition.
Konstantinos RapantzikosYannis AvrithisStefanos D. KolliasPublished in: CVPR (2009)
Keyphrases
- feature points
- action recognition
- image sequences
- bag of words
- human detection
- human actions
- affine transformation
- point correspondences
- epipolar geometry
- pose estimation
- computer vision
- space time
- feature matching
- camera parameters
- spatio temporal
- fundamental matrix
- face images
- bag of features
- moving objects
- facial features
- human faces
- feature descriptors
- saliency map
- human activities
- camera motion
- multi view
- image processing