Exploring Emergent Properties of Recurrent Neural Networks Using a Novel Energy Function Formalism.
Rakesh SenguptaSurampudi Bapi RajuAnindya PattanayakPublished in: LOD (1) (2023)
Keyphrases
- energy function
- recurrent neural networks
- graph cuts
- energy minimization
- markov random field
- global minimum
- neural network
- active contours
- shape model
- image intensity
- segmentation method
- feed forward
- hopfield neural network
- artificial neural networks
- belief propagation
- echo state networks
- image segmentation
- level set
- recurrent networks
- energy functional
- neural model
- reservoir computing
- cascade correlation
- problems in computer vision
- smoothness constraint
- shape prior
- regularization term
- maximum a posteriori
- reconstruction process
- higher order
- chaotic time series
- nonlinear dynamic systems
- probability distribution
- learning algorithm