Mind the Regularized GAP, for Human Action Classification and Semi-supervised Localization based on Visual Saliency.
Marc MoreauxNatalia LyubovaIsabelle FerranéFrédéric LeraslePublished in: VISIGRAPP (4: VISAPP) (2018)
Keyphrases
- visual saliency
- action classification
- semi supervised
- action recognition
- human detection
- saliency map
- natural images
- visual attention
- human visual system
- eye movements
- human actions
- bayesian framework
- bag of words
- computer vision
- activity recognition
- human activities
- higher level
- image classification
- multi class
- active learning
- feature vectors
- natural scenes
- multiscale
- high level
- image processing