Amortized Inference for Heterogeneous Reconstruction in Cryo-EM.
Axel LevyGordon WetzsteinJulien N. P. MartelFrédéric PoitevinEllen D. ZhongPublished in: NeurIPS (2022)
Keyphrases
- expectation maximization
- gibbs sampling
- three dimensional
- bayesian networks
- em algorithm
- compressed sensing
- parameter learning
- high resolution
- image reconstruction
- mixture model
- probabilistic inference
- k means
- external memory
- inference process
- belief networks
- worst case
- search tree
- reconstruction process
- bayesian inference
- reconstruction error
- tomographic reconstruction
- running times
- neural network
- compressive sensing
- heterogeneous data
- gaussian distribution
- gaussian mixture model
- maximum likelihood
- data structure
- real world