Usage Examples¶
Example Dependencies¶
Some examples use additional dependencies, which are listed in examples_requirements.txt.
The additional requirements should be installed via pip, with the exception of astra-toolbox
,
which should be installed via conda:
conda install astra-toolbox
pip install -r examples/examples_requirements.txt # Installs other example requirements
The dependencies can also be installed individually as required.
Note that astra-toolbox
should be installed on a host with one or more CUDA GPUs to ensure
that the version with GPU support is installed.
Run Time¶
Most of these examples have been constructed with sufficiently small test problems to allow them to run to completion within 5 minutes or less on a reasonable workstation. Note, however, that it was not feasible to construct meaningful examples of the training of some of the deep learning algorithms that complete within a relatively short time; the examples “CT Training and Reconstructions with MoDL” and “CT Training and Reconstructions with ODP” in particular are much slower, and can require multiple hours to run on a workstation with multiple GPUs.
Organized by Application¶
Computed Tomography¶
- TV-Regularized Abel Inversion
- Parameter Tuning for TV-Regularized Abel Inversion
- CT Reconstruction with CG and PCG
- 3D TV-Regularized Sparse-View CT Reconstruction (ADMM Solver)
- 3D TV-Regularized Sparse-View CT Reconstruction (Proximal ADMM Solver)
- TV-Regularized Sparse-View CT Reconstruction (ASTRA Projector)
- TV-Regularized Sparse-View CT Reconstruction (Integrated Projector)
- TV-Regularized Low-Dose CT Reconstruction
- TV-Regularized CT Reconstruction (Multiple Algorithms)
- PPP (with BM3D) CT Reconstruction (ADMM with CG Subproblem Solver)
- PPP (with BM3D) CT Reconstruction (ADMM with Fast SVMBIR Prox)
- PPP (with BM3D) Fan-Beam CT Reconstruction
- CT Training and Reconstructions with MoDL
- CT Training and Reconstructions with ODP
- CT Training and Reconstructions with UNet
- X-ray Transform Comparison
- TV-Regularized Sparse-View CT Reconstruction (Multiple Projectors, Common Sinogram)
- TV-Regularized Sparse-View CT Reconstruction (Multiple Projectors)
Deconvolution¶
- Circulant Blur Image Deconvolution with TV Regularization
- Image Deconvolution with TV Regularization (ADMM Solver)
- Image Deconvolution with TV Regularization (Proximal ADMM Solver)
- Parameter Tuning for Image Deconvolution with TV Regularization (ADMM Solver)
- Deconvolution Microscopy (Single Channel)
- Deconvolution Microscopy (All Channels)
- PPP (with BM3D) Image Deconvolution (ADMM Solver)
- PPP (with BM3D) Image Deconvolution (APGM Solver)
- PPP (with DnCNN) Image Deconvolution (ADMM Solver)
- PPP (with DnCNN) Image Deconvolution (Proximal ADMM Solver)
- PPP (with BM4D) Volume Deconvolution
- Deconvolution Training and Reconstructions with MoDL
- Deconvolution Training and Reconstructions with ODP
Sparse Coding¶
Miscellaneous¶
- PPP (with BM3D) Image Demosaicing
- PPP (with DnCNN) Image Superresolution
- ℓ1 Total Variation Denoising
- Total Variation Denoising (ADMM)
- Total Variation Denoising with Constraint (APGM)
- Comparison of Optimization Algorithms for Total Variation Denoising
- Denoising with Approximate Total Variation Proximal Operator
- Complex Total Variation Denoising with NLPADMM Solver
- Complex Total Variation Denoising with PDHG Solver
- Comparison of DnCNN Variants for Image Denoising
- TV-Regularized 3D DiffuserCam Reconstruction
- Video Decomposition via Robust PCA
- CT Data Generation for NN Training
- Blurred Data Generation (Natural Images) for NN Training
- Blurred Data Generation (Foams) for NN Training
- Noisy Data Generation for NN Training
Organized by Regularization¶
Plug and Play Priors¶
- PPP (with BM3D) CT Reconstruction (ADMM with CG Subproblem Solver)
- PPP (with BM3D) CT Reconstruction (ADMM with Fast SVMBIR Prox)
- PPP (with BM3D) Fan-Beam CT Reconstruction
- PPP (with BM3D) Image Deconvolution (ADMM Solver)
- PPP (with BM3D) Image Deconvolution (APGM Solver)
- PPP (with DnCNN) Image Deconvolution (ADMM Solver)
- PPP (with DnCNN) Image Deconvolution (Proximal ADMM Solver)
- PPP (with BM4D) Volume Deconvolution
- PPP (with BM3D) Image Demosaicing
- PPP (with DnCNN) Image Superresolution
Total Variation¶
- TV-Regularized Abel Inversion
- Parameter Tuning for TV-Regularized Abel Inversion
- TV-Regularized Sparse-View CT Reconstruction (ASTRA Projector)
- TV-Regularized Sparse-View CT Reconstruction (Integrated Projector)
- 3D TV-Regularized Sparse-View CT Reconstruction (ADMM Solver)
- 3D TV-Regularized Sparse-View CT Reconstruction (Proximal ADMM Solver)
- TV-Regularized Low-Dose CT Reconstruction
- TV-Regularized CT Reconstruction (Multiple Algorithms)
- Circulant Blur Image Deconvolution with TV Regularization
- Image Deconvolution with TV Regularization (ADMM Solver)
- Parameter Tuning for Image Deconvolution with TV Regularization (ADMM Solver)
- Image Deconvolution with TV Regularization (Proximal ADMM Solver)
- Deconvolution Microscopy (Single Channel)
- Deconvolution Microscopy (All Channels)
- ℓ1 Total Variation Denoising
- Total Variation Denoising (ADMM)
- Total Variation Denoising with Constraint (APGM)
- Comparison of Optimization Algorithms for Total Variation Denoising
- Denoising with Approximate Total Variation Proximal Operator
- Complex Total Variation Denoising with NLPADMM Solver
- Complex Total Variation Denoising with PDHG Solver
- TV-Regularized 3D DiffuserCam Reconstruction
Sparsity¶
- TV-Regularized 3D DiffuserCam Reconstruction
- Non-Negative Basis Pursuit DeNoising (ADMM)
- Non-Negative Basis Pursuit DeNoising (APGM)
- Convolutional Sparse Coding (ADMM)
- Convolutional Sparse Coding with Mask Decoupling (ADMM)
- Basis Pursuit DeNoising (APGM)
- Non-negative Poisson Loss Reconstruction (APGM)
- Video Decomposition via Robust PCA
Machine Learning¶
- CT Data Generation for NN Training
- CT Training and Reconstructions with MoDL
- CT Training and Reconstructions with ODP
- CT Training and Reconstructions with UNet
- Blurred Data Generation (Natural Images) for NN Training
- Blurred Data Generation (Foams) for NN Training
- Deconvolution Training and Reconstructions with MoDL
- Deconvolution Training and Reconstructions with ODP
- Noisy Data Generation for NN Training
- Training of DnCNN for Denoising
- Comparison of DnCNN Variants for Image Denoising
Organized by Optimization Algorithm¶
ADMM¶
- TV-Regularized Abel Inversion
- Parameter Tuning for TV-Regularized Abel Inversion
- TV-Regularized Sparse-View CT Reconstruction (ASTRA Projector)
- TV-Regularized Sparse-View CT Reconstruction (Integrated Projector)
- 3D TV-Regularized Sparse-View CT Reconstruction (ADMM Solver)
- TV-Regularized Low-Dose CT Reconstruction
- TV-Regularized CT Reconstruction (Multiple Algorithms)
- PPP (with BM3D) CT Reconstruction (ADMM with CG Subproblem Solver)
- PPP (with BM3D) CT Reconstruction (ADMM with Fast SVMBIR Prox)
- PPP (with BM3D) Fan-Beam CT Reconstruction
- Circulant Blur Image Deconvolution with TV Regularization
- Image Deconvolution with TV Regularization (ADMM Solver)
- Parameter Tuning for Image Deconvolution with TV Regularization (ADMM Solver)
- Deconvolution Microscopy (Single Channel)
- Deconvolution Microscopy (All Channels)
- PPP (with BM3D) Image Deconvolution (ADMM Solver)
- PPP (with DnCNN) Image Deconvolution (ADMM Solver)
- PPP (with BM4D) Volume Deconvolution
- TV-Regularized 3D DiffuserCam Reconstruction
- Non-Negative Basis Pursuit DeNoising (ADMM)
- Convolutional Sparse Coding (ADMM)
- Convolutional Sparse Coding with Mask Decoupling (ADMM)
- PPP (with BM3D) Image Demosaicing
- PPP (with DnCNN) Image Superresolution
- ℓ1 Total Variation Denoising
- Total Variation Denoising (ADMM)
- Comparison of Optimization Algorithms for Total Variation Denoising
- Denoising with Approximate Total Variation Proximal Operator
- Video Decomposition via Robust PCA