View-Invariant Action Recognition Using Latent Kernelized Structural SVM.
Xinxiao WuYunde JiaPublished in: ECCV (5) (2012)
Keyphrases
- structural svm
- view invariant
- action recognition
- human actions
- latent variables
- multi view
- single view
- discriminative training
- cutting plane algorithm
- quadratic programming
- contextual features
- ls svm
- cutting plane
- co training
- human motion
- spatio temporal
- viewpoint
- probabilistic model
- multiple viewpoints
- probabilistic interpretation
- gesture recognition
- maximum margin
- invariant representation
- activity recognition
- conditional random fields
- least squares
- hidden markov models
- prior knowledge
- three dimensional