nodes/wTechArtist-ComfyUI_VVL_VideoCamera_Advanced
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
65 Commits 3 Branches 0 Tags
main
Commit Graph
65 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
weiweiliang f120e9e966 调整箭头长度显示：增加基准长度到15%，范围8%-20%，提升可视化效果 2025-05-29 21:08:21 +08:00
weiweiliang 6d70090081 修复VGGT箭头长度问题: 使用相对长度避免箭头射出画面外 2025-05-29 20:33:52 +08:00
weiweiliang e1f9d3aa65 更新VGGT节点: 使用3D立体轨迹可视化，与COLMAP节点保持一致的视觉效果 2025-05-29 20:26:38 +08:00
weiweiliang ae725d04bd Clean up: Remove unnecessary files, keep only ComfyUI node essentials 2025-05-29 18:34:22 +08:00
wTechArtist ac738c88bd Initial commit 2025-05-29 18:26:23 +08:00
Jianyuan Wang 22d5c18fe6 [WIP] Upload demo_colmap.py, which can (a) run both vggt and vggt+ba, (b) output in a colmap format, (c) in the original image resolution. The output can be directly used for nerf/gaussian training. 2025-05-22 23:27:58 +01:00
Jianyuan Wang a9f0fa7d6b [WIP] Upload demo_colmap.py, which can (a) run both vggt and vggt+ba, (b) output in a colmap format, (c) in the original image resolution. The output can be directly used for nerf/gaussian training. 2025-05-22 23:24:25 +01:00
Jianyuan Wang c4b5da2d85 Merge pull request #125 from D31T4/main 
Fix vggt_to_colmap images.bin output format
 2025-05-08 11:22:48 +01:00
Jianyuan Wang 3d56b6b6bc Avoid using np allow_pickle for Deserialization 2025-05-07 16:28:25 +01:00
D31T4 3fc15eadda fix vggt_to_colmap images.bin output format 2025-05-07 14:06:24 +10:00
jianyuan b017d60f32 upload eval code for co3d 2025-05-03 16:38:45 +00:00
Jianyuan Wang 51ee022e3d Merge pull request #112 from Vincentqyw/main 
Fix MPS compatibility by using float32 in positional embeddings
 2025-04-28 02:20:27 +01:00
Realcat 65c1f4ed54 fix: use float32 in pos embed for MPS compatibility 2025-04-24 17:41:34 +00:00
Realcat 350abfcdc6 fix: use float32 in pos embed for MPS compatibility 2025-04-24 17:36:37 +00:00
Jianyuan Wang 588a0a238d update the news about training code 2025-04-13 05:19:31 +01:00
Jianyuan Wang 7260ede3ff Merge pull request #86 from djecoo/python_version_uv 
Update uv python version
 2025-04-11 00:37:47 +01:00
djecoo 35e65de083 Update uv python version 2025-04-06 11:02:54 +02:00
Jianyuan Wang b02cc03cee VGGT to COLMAP format converter (Merge pull request #42 from atakan-topaloglu/main) 
Add colmap_writer.py: VGGT to COLMAP format converter
 2025-03-30 04:31:51 +01:00
Atakan Topaloglu 81ddb97ad4 Fix: Removed old colmap_writer.py 2025-03-30 01:46:49 +03:00
Atakan Topaloglu 30180e6e71 Fix: Works for arbitrary captures (not only orbital captures) with bin support, faster and memory efficient point cloud storage, sky and background filtering 2025-03-29 18:50:53 +03:00
Atakan Topaloglu 5c91b0169e Sky filtering works, remove the temp directories and save the final model 2025-03-29 18:12:24 +03:00
Atakan Topaloglu 009d8515d1 Last commit beofre following the convention of visual_util.py 2025-03-29 17:34:34 +03:00
Atakan Topaloglu daf7f94ba4 Sky filtering, and bw filtering works, regression tests run. Will try making confidence filtering identical to that of VGGT 2025-03-29 16:49:25 +03:00
Atakan Topaloglu 3cf03416c3 Color works, smoke test passed, add filter background and sky options. Todo: better confidence thresholding 2025-03-29 16:28:51 +03:00
Atakan Topaloglu 8aa0832859 Fix the incorrect camera poses and added .bin COLMAP exports, point cloud still in grayscale 2025-03-29 16:18:39 +03:00
Jianyuan Wang 6d361a374e Merge pull request #57 from yashbhalgat/yash-preproc-fix 
Add 'pad' mode in preprocessing to handle cases where height > width
 2025-03-25 23:07:46 +00:00
Jianyuan Wang b4b73db60d Merge pull request #55 from pablovela5620/main 
Switching build system broke installation
 2025-03-25 22:59:40 +00:00
yashbhalgat 692096e818 Add 'pad' mode in preprocessing to handle cases where height is greater than width 2025-03-25 13:05:23 +00:00
Pablo Vela 467690acac Update pyproject.toml 2025-03-24 14:57:00 -05:00
Pablo Vela 1a14b2d0e2 Update pyproject.toml 2025-03-24 14:52:22 -05:00
Jianyuan Wang 6830466c42 Update pyproject file to allow all numpy<2 2025-03-23 21:50:05 +00:00
jianyuan b66eeb36b5 provide details to install vggt as a package 2025-03-23 20:25:58 +00:00
jianyuan 16d54825a5 provide details to install vggt as a package 2025-03-23 20:24:54 +00:00
jianyuan e35b89ac6a provide details to install vggt as a package 2025-03-23 20:22:43 +00:00
jianyuan 01e5d11e2a update pyproject and readme 2025-03-23 20:14:54 +00:00
jianyuan 4d9710fb2e update pyproject and readme 2025-03-23 20:07:49 +00:00
Jianyuan Wang 4e321b8a8c Add pyproject.toml to allow easy installion 
Add a proper pyproject.toml to allow easy installion
 2025-03-23 19:51:17 +00:00
pablovela5620 f8bf73ac25 update numpy version constraint to allow for compatibility with future releases 2025-03-23 14:28:55 -05:00
pablovela5620 a6ba138800 relax torch + torchvision version requirements 2025-03-23 14:27:57 -05:00
pablovela5620 d33e01ee13 update installation instructions in README.md for pip, pixi, and uv 2025-03-23 14:22:52 -05:00
pablovela5620 ce841d7dc0 add proper pyproject.toml 2025-03-23 14:12:42 -05:00
Jianyuan Wang a16b0760f6 add instruction about masking 2025-03-22 18:09:30 +00:00
Atakan Topaloglu a902743b97 Changed gitignore for compatibility with main branch 2025-03-20 21:29:17 +03:00
Atakan Topaloglu 74213fcd52 Added confidence thresholding to colmap_writer.py 2025-03-20 21:28:06 +03:00
Atakan Topaloglu 0d31386f53 Remove file from remote 2025-03-20 21:10:58 +03:00
Atakan Topaloglu 67b02db572 Added colmap writer for point cloud and camera initialization 2025-03-20 21:08:48 +03:00
Atakan Topaloglu 3fba88d5bf Merge remote-tracking branch 'upstream/main' 2025-03-20 20:55:36 +03:00
Jianyuan Wang ae1b9c8750 update tensor contiguous for dpt head 2025-03-20 17:42:14 +00:00
Atakan Topaloglu 4b14330514 Point cloud and camera estimation works well 2025-03-20 20:08:51 +03:00
Jianyuan Wang 497e1bade4 add single view results from Karim Abou Zeid 2025-03-19 22:36:04 +00:00