Version 1.2 (Cupy and Meshes)

This is a major update to DeepDRR with multiple high-level changes:

• We have added support for combined mesh- and volume-based rendering of X-rays. This allows for massive large-scale simulation of datasets with very efficient projection of many ground-truth organs, as in this paper.
• We have removed pycuda dependencies in favor of cupy, which is compatible with CUDA 12.
• We have moved the geo package to an independent package, killeengeo, which allows for the geometry utilities to be used without installing all of DeepDRR.
• A SimpleDevice class is added for easier control of a virtual C-arm in simulation.
• Other bug fixes and quality-of-life changes.

Other changes

• Fix case with no air in volume but att outside by @liamjwang in #96
• Switch from pycuda to cupy by @liamjwang in #97
• Improved installation guide and dockerfile by @liamjwang in #100
• Hybrid mesh-voxel DRRs by @liamjwang in #104
• Add API for rendering segmentations, depth hit lists, and hide/show for mesh renderables by @liamjwang in #113
• Add seg batching by @liamjwang in #114
• Batch api depth fix by @liamjwang in #115
• 2x faster depth peeling by @liamjwang in #116
• Multi mat stl by @liamjwang in #117
• Add project distance api by @liamjwang in #118
• Rename by @liamjwang in #119
• Clamp positive by @liamjwang in #122
• Add volume shape assert by @liamjwang in #123
• Add new device, fix intrinsic bug by @liamjwang in #125
• Fix renderable center_in_world by @liamjwang in #127
• Add cap blocks, masks by @liamjwang in #126
• Volume segmentation utils by @liamjwang in #128
• General Custom Material Support by @florencekl in #136
• Fix ASCII encoding error in material by @florencekl in #137
• Version 1.2 Milestone by @benjamindkilleen in #144

New Contributors

• @florencekl made their first contribution in #136

Full Changelog: v1.1.3...v1.2.0

July Milestone Release

What's Changed

• Speed up mesh voxelizer by @liamjwang in #84
• Dev by @benjamindkilleen in #93

New Contributors

• @liamjwang made their first contribution in #84

Full Changelog: v1.1.2...v1.1.3

March 2023 Milestone Release

Changes through March 2023.

What's Changed

• Added support for voxelized surgical instruments from STL files.

Full Changelog: v1.1.1...v1.1.2

August 2022 Milestone Release

Features and improvements through August 2022.

What's Changed

• Faster mass attenuation using CUDA not pycuda, ~20% speedup by @benjamindkilleen in #23
• Remove parentheses from Macro by @maxrohleder in #29
• vol.load_dicom() implementation by @maxrohleder in #30
• V1.1.0 alpha by @benjamindkilleen in #34
• added docs requirements by @benjamindkilleen in #35
• Dev by @benjamindkilleen in #36
• fix requirements by @benjamindkilleen in #37
• re-added static by @benjamindkilleen in #38
• master doc by @benjamindkilleen in #39
• fixed sphinx version by @benjamindkilleen in #40
• path fix by @benjamindkilleen in #41
• Multi volume by @benjamindkilleen in #47
• Dev by @benjamindkilleen in #48
• Mc scatter by @benjamindkilleen in #50
• Slow kernel fix by @benjamindkilleen in #51
• Dev by @benjamindkilleen in #52
• Dev by @benjamindkilleen in #56
• Dev by @benjamindkilleen in #57
• Sean branch by @benjamindkilleen in #61
• Dev by @benjamindkilleen in #62
• Fix indexing in the README by @v715 in #73
• Minor change by @benjamindkilleen in #75
• Release 1.1 by @benjamindkilleen in #76

New Contributors

• @maxrohleder made their first contribution in #29
• @v715 made their first contribution in #73

Full Changelog: v1.0.0...v1.1.1

Stability Improvements

This release contains minor stability improvements and features.

August 2021 Milestone Release

This release contains bug fixes and updates up to August 2021.

Multi-volume Support

This pre-release includes initial support for fast multi-volume rendering.

Version 1.1 test release

Testing PyPI workflow for 1.1 release.

Version 1.0

v1.0.0

fix links

1.0 Pre-release

This release focuses on usability. It is installable on pip.