Justin Kerr
ceb3540b4c
Merge branch 'main' into justin/eval-patch
2024-10-16 12:53:38 -07:00
Justin Kerr
d2f2ad9474
update gsplat version to >=1.3.0 ( #3478 )
...
* update gsplat version
* 1.4.0
2024-10-16 12:32:06 -07:00
Justin Kerr
b046e3b975
add ability to load config.yml from a folder not the same as the original training one
2024-10-16 12:31:49 -07:00
Simon Bethke
adc743c3a0
tag exported ply files ( #3472 )
...
* tag exported ply files
* Added the version
* fix import order
* Added information about the up-axis
2024-10-13 18:48:52 -07:00
Brent Yi
67be06009e
Add option to change .ply output filename for Gaussian exports ( #3467 )
...
Add option to configure .ply output filename for Gaussian exports
2024-10-09 22:22:52 +00:00
Nicholas Pfaff
8e27e5f54d
Add option for refining the TSDF using the mesh's AABB ( #3459 )
2024-10-06 21:23:16 -07:00
Brent Yi
1bc1e1aaf6
Add max_frames + max_output_size arguments for Aria data processing ( #3466 )
...
* Add max_frames + max_image_size arguments for Aria data processing
* Change max_output_size
2024-10-06 21:22:16 -07:00
Nicholas Pfaff
866a069777
Keep depth_image on CPU ( #3449 )
...
Co-authored-by: J.Y. <132313008+jb-ye@users.noreply.github.com>
2024-10-01 19:39:26 +00:00
Nicholas Pfaff
334a1c8afe
Make PixelSampler the depth-nerfacto default ( #3454 )
...
* Make PixelSampler the depth-nerfacto default
* Add note about PairPixelSampler and DepthLossType.SPARSENERF_RANKING
2024-10-01 16:59:31 +00:00
Nicholas Pfaff
56556d22fa
Rejection sample mask for patch pixel sampler ( #3457 )
...
Use mask rejection sampling in PatchPixelSampler
2024-10-01 09:54:28 -07:00
Nicholas Pfaff
4c7297c75a
Speed up PairPixelSampler ( #3452 )
2024-09-30 17:31:49 -07:00
Nicholas Pfaff
6da94443fd
Change depth from float64 to float32 ( #3453 )
2024-09-28 16:51:05 -07:00
J.Y.
01d3f73e2c
Fix #3417 ( #3441 )
2024-09-24 18:01:43 -07:00
Aayush Gupta
e6941aca47
Zipnerf point cloud and tsdf mesh export support ( #3429 )
...
Co-authored-by: J.Y. <132313008+jb-ye@users.noreply.github.com>
2024-09-24 15:28:18 -04:00
Aayush Gupta
516fd7c9ac
Fix splatfacto crash in eval when using viewer ( #3430 )
2024-09-17 12:31:42 -07:00
Anthony Tafoya
3e638da635
Adding Random Seed for Frame Processing ( #3416 )
...
* Adding Random Seed for Frame Processing
* Added Unit Tests
* Updating Unit Tests for Ffmpeg
* Make Logs More Detailed
---------
Co-authored-by: Anthony-Tafoya <anthonytafoya@berkeley.edu>
Co-authored-by: J.Y. <132313008+jb-ye@users.noreply.github.com>
2024-09-17 15:32:09 +00:00
Nicholas Pfaff
27b8e14798
Fix crop obb masking in pointcloud generation ( #3433 )
2024-09-16 16:44:31 -04:00
Maximum Wilder-Smith
106d427e80
Minor update to generate splat files in inference mode ( #3432 )
...
changed gaussian splat export to inference mode
2024-09-16 13:57:21 -04:00
J.Y.
7a0b8cbc89
use gsplat strategy interface to simplify splatfacto ( #3376 )
...
* use gsplat strategy interface to simplify splatfacto
* change splatfacto-big densify_grad_thresh to 0.0005
* upgrade gsplat version
* filter low opacities when exporting
* remove num_sh_bases
---------
Co-authored-by: Jianbo Ye <jianboye@amazon.com>
Co-authored-by: Justin Kerr <justin.g.kerr@gmail.com>
2024-09-13 07:10:37 -04:00
Jonáš Kulhánek
194b5d4e46
Fix docker attestation fail on forked repo PRs ( #3423 )
2024-09-12 09:43:36 -04:00
Aayush Gupta
e7b7dc9c6f
Add option to start training paused ( #3420 )
2024-09-10 19:56:44 +00:00
Daniel Crispell
e2a0915e56
bugfix: handle multiple camera types in a single batch ( #3415 )
2024-09-10 15:31:05 -04:00
Hardik Dava
d67b281f7d
RGBA renderer for splatfacto ( #3307 )
...
* Update base_model.py
* Update render.py
* Update render.py
* fixing error
* Update render.py
* Update render.py
2024-09-09 16:20:26 -04:00
Jonáš Kulhánek
f86dbe6d45
Update dockerfile and add docker build action ( #3283 )
...
* Use github docker registry
* Fix DDP train for GPU in exclusive mode
* Improve docker image - compile gsplat, decrease image size
* Drop unrelated change
* Add build docker image action
* Rename build docker image action
* nit
* Remove commented line from Dockerfile
* Fix dockerfile when explicit source is specified
* Fix failing dynamo build for torch.compile
* Lock dockerfile and tcnn versions
* Add `torch.cuda.is_available()` condition
* Drop set_cuda_device
* Fix build docker image github action
* Docker build save disk space
* Set MAX_JOBS to limit resource usage for docker build
* Try bumping `MAX_JOBS` 2 => 4
* Install fixed gsplat version from nerfstudio's pyproject.toml
* Update docs
* Finish docker build action
* Fix ignores push when PR
---------
Co-authored-by: Gina Wu <ginawu98@gmail.com>
Co-authored-by: Brent Yi <yibrenth@gmail.com>
2024-09-05 10:12:21 +02:00
Hoanh Le
96b7fe2230
update documentation for instant-ngp ( #3406 )
...
update documentation
2024-09-03 10:09:18 -04:00
jiaruiyu99
a6214921f8
Add eval data to image_rename_map ( #3396 )
...
* Add eval data to image_rename_map
* Follow name convention
---------
Co-authored-by: Brent Yi <yibrenth@gmail.com>
Co-authored-by: J.Y. <132313008+jb-ye@users.noreply.github.com>
2024-09-03 01:35:54 +00:00
Hoanh Le
d395aa2e53
change grid resolution type ( #3400 )
2024-08-30 17:32:42 +03:00
Gina Wu
dff42c9a81
Bump viser to 0.2.7 ( #3398 )
...
Bump viser 0.2.7
Co-authored-by: Brent Yi <yibrenth@gmail.com>
2024-08-29 17:53:34 -07:00
Brent Yi
33dce833ee
Suppress wandb type error ( #3397 )
2024-08-29 17:21:45 -07:00
J.Y.
6410080244
bump version to v1.1.4 ( #3394 )
...
bump version
Co-authored-by: Jianbo Ye <jianboye@amazon.com>
2024-08-28 17:06:59 -04:00
J.Y.
96ca8b0ab7
Update exporter.py to export sh_degree 0 case #3371 ( #3374 )
...
* Update exporter.py for sh_degree 0
Change to write sh coefficients instead of color values
* Add flag for use_sh0_renderer
Add sh0 renderer case for model.config.sh_degree == 0
* fix ruff
* add warning if use_sh0_renderer is used when higher order of SH is available
* fix rgb export for color-only training
* use ply_color_mode
* better handling ply_color_mode=='rgb' when sh_degree>0
* clean RGB2SH
* fix issues
* update description
---------
Co-authored-by: bell-one <bellone1991@gmail.com>
Co-authored-by: Jianbo Ye <jianboye@amazon.com>
Co-authored-by: Brent Yi <yibrenth@gmail.com>
2024-08-28 19:41:25 +00:00
Hoanh Le
22dae34eeb
add assertion to make sure positions and positions_flat are not empty ( #3387 )
2024-08-28 12:37:54 -07:00
Anthony Zhang
69ae4abbfd
Updating camera intrinsics (principal points) when the undistorted image is cropped ( #3382 )
...
* added notes and a possible bugfix when region of interest crops image during undistortion to update intrinsics
* fixing index error, principal points are on the 3rd column of 3x3 matrix K
---------
Co-authored-by: Brent Yi <yibrenth@gmail.com>
Co-authored-by: J.Y. <132313008+jb-ye@users.noreply.github.com>
2024-08-26 17:26:51 -07:00
Ruilong Li(李瑞龙)
1f2344a731
get rid of gsplat legacy dep ( #3386 )
...
Co-authored-by: Ruilong Li <397653553@qq.com>
2024-08-26 11:28:34 -04:00
Yufeng Zhang
792c5fa9f7
Add an example on how to customize dataparser with ns-train ( #3385 )
2024-08-25 22:04:23 +00:00
J.Y.
02d45c1fd7
Fix #3355 ( #3381 )
...
* fix nonzero colmap K4
2024-08-25 14:48:17 -07:00
Brent Yi
b707727950
Fix memory leak in full_images_datamanager.py ( #3375 )
...
* Fix memory leak in full_images_datamanager.py
* remove trailing space
* follow-up remove some deepcopy
---------
Co-authored-by: J.Y. <132313008+jb-ye@users.noreply.github.com>
Co-authored-by: Jianbo Ye <jianboye@amazon.com>
2024-08-23 11:04:32 -07:00
pierremerriaux-leddartech
e3a506f346
fix bugs in generfacto, train_dataparser_outputs.save_dataparser_transform and viewer camera on cpu ( #3378 )
...
* fix bugs in generfacto, train_dataparser_outputs.save_dataparser_transform and viewer camera on cpu
* fix formating
2024-08-23 11:44:47 -04:00
Alex Jean
36d5e35801
Addressed TODO in full_images_datamanager.py to return image resolution as train rays ( #3312 )
...
* addressed a todo in full_images_datamanager.py that should return the resolution of the image being rendered
* fixed docstring get_train_rays_per_batch
2024-08-21 21:59:44 +00:00
J.Y.
a281d99eac
bug fix for splatfacto ( #3367 )
...
* bug fix for splatfacto
* fix core tests
* Remove ipynb exclude for ruff, will fix in #3369
---------
Co-authored-by: Brent Yi <yibrenth@gmail.com>
2024-08-19 16:11:12 -04:00
Sebastiaan
096116097a
installation.md: Remove manual gsplat installation for Windows ( #3333 )
...
* installation.md: Remove manual gsplat installation for Windows
* installation.md: Reformat and clarify Windows instructions.
* Update installation.md
installation.md: Minor change in wording
---------
Co-authored-by: J.Y. <132313008+jb-ye@users.noreply.github.com>
2024-08-19 15:44:41 +00:00
Brent Yi
c669d865c1
Bump + run ruff ( #3369 )
...
* Bump + run ruff
* Fix notebook import
2024-08-19 11:23:14 -04:00
Congrong Xu
2ed1716980
Bilagird for splatfacto ( #3316 )
...
add bilateral grid as option to splatfacto training
2024-08-13 19:18:43 +00:00
Ethan Weber
ec10c49d51
Lazy import for open3d in Blender DataParser ( #3347 )
...
lazy import for open3d in blender dataparser
Co-authored-by: Ethan Weber <ethanjohnweber@meta.com>
2024-08-05 18:55:59 +00:00
Congrong Xu
9ffbc374f4
Allows rendering the appearance of specific training camera indices ( #3346 )
...
allows render given camera idx
2024-08-05 10:41:55 -07:00
Guangyun-Xu
b0dd3d4730
make np.array compatible with numpy 2.0 ( #3339 )
...
* make np.array compatible with numpy 2.0
* Update camera_utils.py
2024-07-29 12:44:06 -07:00
Sebastiaan
8c87906971
Swap out opencv-python for opencv-python-headless ( #3336 )
2024-07-27 15:48:56 -04:00
J.Y.
702886b213
fix ruff check ( #3328 )
2024-07-25 09:22:59 +09:00
Congrong Xu
3a46c4274a
Pass current camera idx in metadata to get output from viser ( #3293 )
...
allows viser camera to pass clicked training camera idx into metadata
2024-07-18 08:03:18 +00:00
Congrong Xu
d2510aa02f
Splatw doc updates ( #3311 )
...
* splatw doc
* fix
* fix typo
* update doc
---------
Co-authored-by: Justin Kerr <justin.g.kerr@gmail.com>
2024-07-18 16:59:55 +09:00