Self-Supervised Learning of Physics-Based Reconstruction Neural Networks without Fully-Sampled Reference Data.
Burhaneddin YamanSeyed Amir Hossein HosseiniSteen MoellerJutta EllermannKâmil UgurbilMehmet AkçakayaPublished in: CoRR (2019)
Keyphrases
- data sets
- neural network
- data processing
- data analysis
- database
- original data
- raw data
- data sources
- experimental data
- data collection
- synthetic data
- learning systems
- prior knowledge
- learning process
- training data
- three dimensional
- sensor data
- background knowledge
- missing data
- learning models
- connectionist systems
- active learning
- data mining
- high quality
- supervised learning
- online learning
- knowledge acquisition
- knowledge discovery
- data points
- image reconstruction
- computer systems
- human experts
- labeled data
- feature selection
- feed forward
- recurrent neural networks
- learning algorithm
- hidden variables
- statistical analysis