A constrained formulation for compressive spectral image reconstruction using linear mixture models.
Jorge BaccaHéctor VargasHenry ArguelloPublished in: CAMSAP (2017)
Keyphrases
- image reconstruction
- mixture model
- gaussian mixture model
- random projections
- reconstructed image
- em algorithm
- generative model
- compressed sensing
- probabilistic model
- density estimation
- model selection
- super resolution
- generalized em algorithm
- expectation maximization
- maximum likelihood
- emission computed tomography
- mixture modeling
- unsupervised learning
- maximum a posteriori
- reconstruction method
- compressive sensing
- language model
- probability density function
- limited angle
- emission tomography
- image reconstruction algorithms
- spectral estimation
- active learning
- pattern recognition
- bayesian networks
- computer vision
- information retrieval