Unsupervised learning in second-order neural networks for motion analysis.
Tomás MaulSapiyan BabaPublished in: Neurocomputing (2011)
Keyphrases
- motion analysis
- unsupervised learning
- neural network
- image sequences
- motion tracking
- supervised learning
- motion model
- pattern recognition
- feature tracking
- expectation maximization
- object recognition
- motion parameters
- point correspondences
- feature selection
- flow field
- dimensionality reduction
- semi supervised
- nonrigid motion
- motion patterns
- articulated objects
- model selection
- text classification
- d objects
- artificial neural networks
- multiscale
- training data
- motion features