MoDL: Model Based Deep Learning Architecture for Inverse Problems.
Hemant Kumar AggarwalMerry P. ManiMathews JacobPublished in: CoRR (2017)
Keyphrases
- deep learning
- inverse problems
- image reconstruction
- global optimization
- bayes optimal
- convex optimization
- unsupervised learning
- machine learning
- optimization methods
- optimization problems
- partial differential equations
- discretization method
- early vision
- mental models
- computationally expensive
- weakly supervised
- smoothness constraint
- higher order
- feature vectors
- reinforcement learning
- computer vision