What's Changed

• Fix translation in double geodesic formula by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/310
• Add optional alphaminimax inputs by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/312
• Remove version locks on pytorch / torchio by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/313
• Implement inplace detector rescaling by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/317
• Add intrinsic parameter resampling by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/318
• Make PoseRegressor return a RigidTransform by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/319

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.4.3...v0.4.4

What's Changed

• Update type information in read function by @haarisr in https://github.com/eigenvivek/DiffDRR/pull/287
• Expose persistent flag to DRR module by @haarisr in https://github.com/eigenvivek/DiffDRR/pull/286
• Get voxel coordinates with volume affine by @coolteemf in https://github.com/eigenvivek/DiffDRR/pull/283
• Orient surface mesh using the volume's affine matrix by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/294
• Fix gradients in Siddon's method by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/296
• Sample alphas for Trilinear using alphaminimax by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/297
• v0.4.3 by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/293
• Add missing parameter mode by @haarisr in https://github.com/eigenvivek/DiffDRR/pull/300
• Remove duplicated lines by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/301
• Initialize calibration matrix with density by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/304
• Make extraction of the largest connected component optional by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/305

New Contributors

• @coolteemf made their first contribution in https://github.com/eigenvivek/DiffDRR/pull/283

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.4.2...v0.4.3

What's Changed

• Update type information in read function by @haarisr in https://github.com/eigenvivek/DiffDRR/pull/287
• Expose persistent flag to DRR module by @haarisr in https://github.com/eigenvivek/DiffDRR/pull/286
• Get voxel coordinates with volume affine by @coolteemf in https://github.com/eigenvivek/DiffDRR/pull/283
• Orient surface mesh using the volume's affine matrix by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/294
• Fix gradients in Siddon's method by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/296
• Sample alphas for Trilinear using alphaminimax by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/297
• v0.4.3 by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/293
• Add missing parameter mode by @haarisr in https://github.com/eigenvivek/DiffDRR/pull/300
• Remove duplicated lines by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/301
• Initialize calibration matrix with density by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/304
• Make extraction of the largest connected component optional by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/305
• Bump version: 0.4.2 → 0.4.3 by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/308

New Contributors

• @coolteemf made their first contribution in https://github.com/eigenvivek/DiffDRR/pull/283

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.4.2...v0.4.3

What's Changed

A bunch more improvements to the implementations of diffdrr.renderers.Siddon and diffdrr.renderers.Trilinear. Siddon's rendering time is down to 37.9 ms -> 29.5 ms -> 26.9 ms!

• Make no_grad in Siddon switchable by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/275
• Add registration example with L-BFGS by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/277
• Reimplement alphaminimax filtering by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/279
• Implement alpha filtering in Trilinear by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/281

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.4.1...v0.4.2

What's Changed

• Add missing parameter to set_intrinsic by @haarisr in https://github.com/eigenvivek/DiffDRR/pull/251
• Add reconstruction tutorial by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/256
• Add normalization to reconstruction tutorial by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/257
• Fix outdated API calls by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/263
• Enable rendering from multiple source points by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/260
• Add 3D rendering of registration by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/264
• Expose intrinsic parameters by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/268
• Fix backprojection (drr.inverse_projection) by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/270
• Reimplement Siddon's method with grid_sample by @momen-mh in https://github.com/eigenvivek/DiffDRR/pull/272

New Contributors

• @haarisr made their first contribution in https://github.com/eigenvivek/DiffDRR/pull/251
• @momen-mh made their first contribution in https://github.com/eigenvivek/DiffDRR/pull/272

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.4.0...v0.4.1

What's Changed

• DiffDRR v0.4 by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/207
• Segmentation by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/212
• Render labelmaps with PyVista by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/214
• Implement efficient rendering of segmentation masks by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/216
• Plot masks by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/218
• Implement geodesics on SE(3) by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/221
• Make new Registration submodule by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/222
• Make mask_to_channels compatible with trilinear renderer by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/230
• Implement SVDO+ from Levinson et al., 2020 by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/232
• Fix perspective projection when reverse_x_axis is True by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/238
• Provide simpler access to camera intrinsics for optimization by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/240
• Hotfix docs by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/242
• Update camera conventions by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/245
• Create greetings.yml by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/247
• Reorient fiducials using affine matrix by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/246
• Add geometry tutorial by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/248
• Update files for v0.4.0 release by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/249

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.12...v0.4.0

What's Changed

• Add notebook for trilinear by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/196
• Add perspective projection and inverse by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/195

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.11...v0.3.12

What's Changed

• Hotfix volume permute by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/189

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.10...v0.3.11

What's Changed

• Move dims to device for torch.compile by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/182
• Switch SE(3) backend to diffdrr.pose.RigidTransform by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/185
• Add trilinear interpolation for rendering by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/187

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.9...v0.3.10

What's Changed

• Interactive pyvista plots by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/174
• Add pytorchse3 by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/176
• Support autodifferentiability for intrinsic parameters by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/178
• Remove dependency on pytorch3d by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/180

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.8...v0.3.9

What's Changed

• Add principal point from intrinsic matrix by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/168
• Add 3D visualization example by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/169
• Misc fixes by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/170

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.7...v0.3.8

What's Changed

• Add patch-based variants of image similarity metrics by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/154
• Implement bone attenuation multiplier by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/155
• Allow bone_attenuation_multiplier to be set once by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/156
• Expose InstanceNorm eps argument by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/160

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.6...v0.3.7

What's Changed

• Support every SO(3) parameterization under the sun by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/149
• Use conda for testing and docs deployment by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/150
• Implement 3D visualization functions with pyvista by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/151

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.5...v0.3.6

Small tweaks to the detector (ensured rays only cast from source to target) and add a new image similarity metric.

What's Changed

• Cleanup experiment folder by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/140
• Detector tweaks by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/143
• Implement Gradient NCC by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/146

Full Changelog: https://github.com/eigenvivek/DiffDRR/commits/v0.3.5

Prioritize differentiable rendering

Final API change for DiffDRR

DRR objects now act purely as a differentiable renderer. That means DRR objects take in rotations and translations parameters to specify (batched) camera poses, and render X-rays from those poses.

The pose parameters are not stored as nn.Parameters of the module. This introduced too many annoyances with the DRR module, such as gradients not flowing backwards past the pose parameters, or only being able to render a specific batch size.

2D-to-3D registration is still possible with DiffDRR. This is done by creating a lightweight wrapper module that stores rotations and translations as nn.Parameters, and calls a DRR object in its forward call. An example is available in the updated documentation.

Small changes:

• DRR generation can now be performed patch-wise (i.e., generate a 600^2 DRR as four 300^2 in series). This means DiffDRR can generate insanely large DRRs since computation memory is no longer a constraint. See an example in the updated timing notebook.

What's Changed

• Differentiable rendering and large DRRs by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/139

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.2...v0.3.3

What's Changed

• Enable cuda testing with GitHub CI by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/135
• New unified API for generating DRRs

The DRR module will now behave as follows:

• Initialize the DRR module as before (volume, spacing, height, pixel_spacing) except don't pass any pose parameters. Instead, initialize a set of dummy pose parameters that are internal to the model.
• Rename _update_params to move_carm. Change the function signature to input two batched vectors, one for translations and rotations. This will be the only way to reposition the C-Arm.
• To render a DRR, only use the forward call (i.e., drr()). No separate project function. This will return a batch of DRRs that are differentiable wrt to all downstream parameters.

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.1...v0.3.2

What's Changed

• Add conversion to DeepDRR by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/134

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.3.0...v0.3.1

DiffDRR is now built using nbdev as the backend! Main advantage: there's now a functional docs site: https://vivekg.dev/DiffDRR

What's Changed

• Make default DiffDRR dtype torch.float32 by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/123
• Implement with nbdev by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/126

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.2.3...v0.3.0

What's Changed

• Noooooow batching should fully work by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/119
• Redo Siddon class as a function by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/121

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.2.1...v0.2.3

Main changes

• Batching fully works!
• Output DRRs have shape [b c h w]
• Initialization of parameters are different now
• You can use .to() to move DRR to devices and dtypes

What's Changed

• Update camera.py by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/109
• Adopt PyTorch conventions and best practices by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/116
• Update plotting function by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/118

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.1.5...v0.2.1

What's Changed

• Fix batching and add test data by @eigenvivek in https://github.com/eigenvivek/DiffDRR/pull/108

Full Changelog: https://github.com/eigenvivek/DiffDRR/compare/v0.1.3...v0.1.5