v1.5

What's Changed

• feat: add camera datum augmentations by @chrisochoatri in #126
• feat: add GitHub issue templates with default assignees by @tk-woven in #127
• docs: remove version-hardcofing from Sphinx by @ryo-takahashi-1206 in #128
• fix: do the deep clone in a GH action, "doc-gen" by @ryo-takahashi-1206 in #129
• fix: supported extensions in compute_image_statistics method by @yuta-tsuzuki-woven in #130
• fix: bug in valid_extensions in compute_image_statistics method by @yuta-tsuzuki-woven in #131
• feat: add dgp2wicker conversion by @chrisochoatri in #132
• fix: fixed depth in dgp2wicker by @chrisochoatri in #133
• test: update paths for key_line_2d unit tests by @nehalmamgain in #136
• feat: add key_line_3d and key_point_3d annotation objects and structures by @nehalmamgain in #135

New Contributors

• @ryo-takahashi-1206 made their first contribution in #128

Full Changelog: v1.4...v1.5

v1.4

What's Changed

• fix: virtual environment to be able to run make test by @nehalmamgain in #110
• fix: fix documentation build and docs by @tk-woven in #113
• feat: fix latent pylint errors in many source files by @tk-woven in #115
• feat: introduce pre-commit and configure most existing linters in it by @tk-woven in #118
• feat: move more linters from SuperLinter to pre-commit by @tk-woven in #122
• feat: replace markdownlint with prettier Markdown auto-formatting by @tk-woven in #123
• ci: updating packages to downstream, including Ubuntu20 and py39 by @wadimkehl in #125
• fix: Update requirements.txt by @wadimkehl in #111
• fix: fixing the numpy dtype warnings > v1.19 by @wadimkehl in #117
• feat: replace last SuperLinter check yamllint with pre-commit hook by @tk-woven in #124

New Contributors

• @wadimkehl made their first contribution in #125

Full Changelog: v1.3...v1.4

TRI-ML DGP v1.3

The v1.3 release contains the definition of some 3D types and some tooling bug fixes.

• add instructions for protos contribution by @nehalmamgain in #103
• fix base image name in Dockerfile by @tk-woven in #107
• symlink python3, not python, to python3.7 by @tk-woven in #108
• add KeyPoint3D, KeyLine3D, Polygon3D by @nehalmamgain in #101

Full Changelog: v1.2...v1.3

TRI-ML DGP v1.2

V1.2 release contains some updates to project dependencies and more.

• Bump ujson from 1.35 to 5.2.0.
• Add support for additional camera models in camera intrinsics.
• Add distortion parameters to BaseDataset.
• Split requirements into lib and dev.
• Add option to specify num_processes in compute_image_statistics.

TRI-ML DGP v1.1

V1.1 release contains the relevant schemas and APIs to expose TRI's autonomous vehicle dataset in a researcher-friendly manner.

• Support ParallelDomainSampleMetadata.
• Support keyline2d and classification annotation schema.
• Support map proto schema.
• Improve test coverage.

TRI-ML DGP v1.0

V1.0 release contains the relevant schemas and APIs to expose autonomous vehicle datasets in a researcher-friendly manner.

• SceneDataset schema defined in protocol buffers (added radar message).
• PyTorch Dataloaders to load DGP datasets.
• CLI for handling and visualizing DGP datasets.

TRI-ML DGP v0.1

This release contains the relevant schemas and APIs to expose TRI's autonomous vehicle dataset in a researcher-friendly manner.

• DGP SceneDataset schema defined in protocol buffers.
• SynchronizedDataset/ParallelDomainDataset to
• load DGP datasets.
• Main CLI for handling DGP datasets.