Generating Human Action Videos by Coupling 3D Game Engines and Probabilistic Graphical Models.
César Roberto de SouzaAdrien GaidonYohann CabonNaila MurrayAntonio M. LópezPublished in: Int. J. Comput. Vis. (2020)
Keyphrases
- human actions
- probabilistic graphical models
- action recognition
- graphical models
- spatio temporal
- human activities
- video sequences
- space time
- human motion
- markov networks
- first order logic
- latent variables
- activity recognition
- action classification
- visual features
- hidden variables
- view invariant
- belief propagation
- approximate inference
- recognizing actions
- probabilistic inference
- parameter learning
- conditional random fields
- random variables
- exact inference
- probabilistic model
- motion recognition
- action recognition in videos
- maximum likelihood
- maximum margin
- fuzzy measures
- multi view
- hidden markov models
- feature extraction
- image sequences
- computer vision
- machine learning