v0.1.8

TLDR

• Record3D users should use the ns-process-data script to convert their data to nerfstudio-data. The record3d-data option in ns-train will be deprecated.
• Mostly Bugfixes

What's Changed

• Fix ffmpeg call by @tancik in #856
• Fix camera interpolation bug by @tancik in #858
• Add colab notebook to repo by @tancik in #864
• Viewer visualization crashes when using DDP by @thomasw21 in #863
• Update colab links to use repo colab by @tancik in #866
• Add more debug logs to data processing by @tancik in #867
• Update readme by @tancik in #869
• Don't add relative paths to experiment name by @tancik in #871
• Add control for resampling images by @kevinddchen in #847
• always list paths as posix paths by @terrancewang in #860
• Check main thread when always rendering. by @thomasw21 in #865
• Add median depth calculation by @tancik in #877
• remove sf street by @terrancewang in #876
• disallow fars <= nears by @decrispell in #881
• add correct none check to distortion params by @nikmo33 in #878
• DNeRF Data Parser by @JulianKnodt in #817
• tensorf by @terrancewang in #586
• Update render UI by @tancik in #885
• [WIP] Record3D Dataparser moved to ns-process-data script. by @akristoffersen in #872
• v0.1.8 by @tancik in #890

New Contributors

• @thomasw21 made their first contribution in #863
• @decrispell made their first contribution in #881
• @nikmo33 made their first contribution in #878

Full Changelog: v0.1.7...v0.1.8

v0.1.7

TLDR

• Mostly bug fixes

What's Changed

• Add D-NeRF dataset download by @JulianKnodt in #815
• Fix pose translation when orientation_method=none by @kevinddchen in #823
• removing quotes on ffmpeg command by @terrancewang in #822
• Fix completion install script for bash by @brentyi in #830
• Pathlib Python 3.7 Compatibility Fix by @akristoffersen in #831
• Normalize depth for packed samples by @kevinddchen in #824
• Fix skipping non-existent images by @kevinddchen in #832
• Fix semantic NeRF config bug by @dolhasz in #838
• docs(scripts/eval.py) Update comments by @ArjunSharda in #841
• Terrance/process data fix by @terrancewang in #842
• Limit number of images in viewer by @kevinddchen in #845
• New Colab Added to README and Docs by @akristoffersen in #848
• Hotfix: add button to master readme by @akristoffersen in #849
• Add github pypi publish action by @tancik in #850
• Add option to render frames by @tancik in #851
• v0.1.7 by @tancik in #852

New Contributors

• @JulianKnodt made their first contribution in #815
• @kevinddchen made their first contribution in #823
• @dolhasz made their first contribution in #838
• @ArjunSharda made their first contribution in #841

Full Changelog: v0.1.6...v0.1.7

v0.1.6

TLDR

• Now supports python 3.7
• Minor bug fixes and speedups ~1.1x

What's Changed

• General acceleration of nerfacto by @liruilong940607 in #797
• Add more controls for orienting poses by @tancik in #800
• Support python 3.7 by @tancik in #802
• Add ability to load path from file by @origamiman72 in #804
• Add viewer version function by @ethanweber in #801
• Update viewer UI by @tancik in #808
• Multi-gpu improvements by @ethanweber in #799
• Fix camera rounding bug by @tancik in #816
• v0.1.6 by @tancik in #820

Full Changelog: v0.1.5.1...v0.1.6

v0.1.5.1

Highlights

• Mostly bug fixes

What's Changed

• Color outside of render view by @tancik in #790
• Update insta360 docs by @tancik in #793
• [VIEWER] Camera Select Bug Fixes by @mcallisterdavid in #791
• Quick fix to zmq ip address by @ethanweber in #795
• Speedup get_ray_sample by @tancik in #794
• v0.1.5.1 by @tancik in #796

Full Changelog: v0.1.5...v0.1.5.1

v0.1.5

Highlights

• Speedup of all methods. nerfacto method is now 1.4x faster.
• Support insta360 data
• Adjust COLMAP parameters for sharper reconstructions (requires reprocessing dataset)

CLI Change

The ns-process-data now requires one to specify the dataset type, current options are [images,video,insta360]
ie.

ns-process-data {images,video,insta360} --data {DATA} --output-dir {OUTPUT_DIR}

Sharper results

Before:

After:

What's Changed

• Fix tb and wandb crash by @tancik in #769
• Snap to cameras by @origamiman72 in #550
• Update first_nerf.md by @krahets in #773
• Update prettier permissions by @tancik in #774
• Add PyTorch jit-ing to the undistort computation for speed by @kerrj in #776
• Sample proposal networks on a schedule to speed training by @kerrj in #770
• Google Colab support with viewer by @ethanweber in #747
• Respect custom timestamps passed to ns-train by @corbt in #778
• Fix tab completion for PyPI installs, add py.typed by @brentyi in #777
• Update custom_dataset.md by @krahets in #771
• Fix threading bug by @tancik in #780
• Add support for insta360 data by @tancik in #781
• Remove prettier workflow by @tancik in #782
• Quick fix to ZMQ deadlock error by @ethanweber in #784
• Refine intrinsics with COLMAP by @tancik in #785
• v0.1.5 by @tancik in #786

New Contributors

• @krahets made their first contribution in #773
• @kerrj made their first contribution in #776
• @corbt made their first contribution in #778

Full Changelog: v0.1.4...v0.1.5

v0.1.4

Highlights

• Reduce nerfacto artifacts where bands would appear across video and objects would disappear.
• Improve nerfacto memory usage
• Multiple bug fixes around dataprocessing and saving and loading checkpoints

Before:

After:

What's Changed

• Fix view_dataset.py by @brentyi in #626
• adding keyboard camera control instructions to viewer by @terrancewang in #735
• bump nerfacc to 0.2.0. bug fix for depth vis. by @liruilong940607 in #744
• fixed --start-train flag by @evonneng in #741
• Added up vector offset by @origamiman72 in #749
• Link fix in README by @akristoffersen in #752
• Improve process data with filename filters and saving mp4s by @ethanweber in #756
• Improving checkpoint logic by @ethanweber in #757
• Reducing flicker artifacts for nerfacto method by @ethanweber in #750
• Add script filenames to CLI docs by @ethanweber in #758
• Bug fixes on Instant NGP dataparser by @akristoffersen in #760
• Reduce memory usage of nerfacto model by @tancik in #766
• v0.1.4 by @tancik in #767

Full Changelog: v0.1.3...v0.1.4

v0.1.3

TLDR

• quick fix to turn off camera pose optimization for non-nerfacto methods
• save renders to renders/ folder instead of output.mp4 by default

What's Changed

• add video link to record3d by @evonneng in #731
• update camera opt defaults by @ethanweber in #732
• save renders to a folder and bump viewer version by @ethanweber in #733
• bump version by @ethanweber in #736

Full Changelog: v0.1.2...v0.1.3

v0.1.2

TLDR

• Fix issue where checkpoints are deleted at the end of training.
• Enable camera pose optimization by default.
• Document record3d

What's Changed

• Record3D docs added to Custom Data docs by @akristoffersen in #718
• Update custom_dataset.md by @akanazawa in #723
• Fix typos by @erich666 in #712
• integrating record3d and fix checkpoint delete issue by @ethanweber in #722
• Pose optimization by @ethanweber in #721
• Pose Opt Doc typo by @akristoffersen in #726
• turn on camera opt by default by @ethanweber in #727
• fix optimizer dict by @ethanweber in #729
• v0.1.2 by @ethanweber in #728

New Contributors

• @erich666 made their first contribution in #712

Full Changelog: v0.1.1...v0.1.2

v0.1.1

TLDR

• Clean up logging, add more error messages
• Update nerfacto defaults
• Clean up Docs

What's Changed

• Add gifs to docs by @tancik in #687
• adding viewer video + other doc related updates by @evonneng in #688
• Update README.md by @akanazawa in #690
• Fix typo in readme by @rohitdatta in #691
• README typos by @darshanthaker in #692
• Installation Docs Updates by @ravirahman in #671
• typo fix (missing space) by @benji-york in #694
• Fix viewer slider stepping past min/max limits by @brentyi in #693
• Readme Changes by @Averylamp in #636
• Update colmap install guide by @tancik in #696
• Remove outdated viewer logging by @tancik in #697
• Fix typos in README by @eleyng in #700
• semantic nerfw docs update by @ethanweber in #704
• Rename: dcargs -> tyro by @brentyi in #702
• Fix typos in index.md by @sean-public in #708
• Update README.md by @mcallisterdavid in #710
• Check ffmpeg before rendering video by @tancik in #715
• Don't apply near plane during eval by @tancik in #716
• max number of iterations logic by @ethanweber in #699
• Reduced tyro overhead + helptext formatting by @brentyi in #719
• v0.1.1 by @tancik in #720

New Contributors

• @rohitdatta made their first contribution in #691
• @darshanthaker made their first contribution in #692
• @ravirahman made their first contribution in #671
• @benji-york made their first contribution in #694
• @Averylamp made their first contribution in #636
• @eleyng made their first contribution in #700
• @sean-public made their first contribution in #708

Full Changelog: v0.1.0...v0.1.1

v0.1.0

What's Changed

Everything

New Contributors

• @tancik made their first contribution in #1
• @evonneng made their first contribution in #3
• @ethanweber made their first contribution in #8
• @brentyi made their first contribution in #105
• @liruilong940607 made their first contribution in #116
• @akristoffersen made their first contribution in #181
• @origamiman72 made their first contribution in #190
• @jake-austin made their first contribution in #168
• @terrancewang made their first contribution in #218
• @mcallisterdavid made their first contribution in #531
• @akanazawa made their first contribution in #653

Full Changelog: https://github.com/nerfstudio-project/nerfstudio/commits/v0.1.0