Recovery Guarantees of Unsupervised Neural Networks for Inverse Problems trained with Gradient Descent.
Nathan BuskulicJalal FadiliYvain QuéauPublished in: CoRR (2024)
Keyphrases
- inverse problems
- neural network
- multilayer perceptron
- image reconstruction
- global optimization
- convex optimization
- training process
- optimization methods
- pattern recognition
- back propagation
- cost function
- optimization problems
- emission tomography
- artificial neural networks
- loss function
- objective function
- early vision
- genetic algorithm
- semi supervised
- partial differential equations
- multiscale
- super resolution
- computationally expensive
- natural images
- simulated annealing
- parameter space
- training set
- feature selection