Downloads · 30 days
0
victor678/pointgt
pointgt is a machine learning model from victor678. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
Artifacts for PointGT: Simultaneous Geometry and Texture Editing for Point-Based Representations (ECCV 2026). Code: <https://github.com/zvict/pointgt
Downloads · 30 days
0
Access
Public
Updated Sep 4, 2026
Repo size
517 MB
Likes
0
Public
Click a slice to open those files.
.pth470 MB · 91%
From the Hugging Face model README
Artifacts for PointGT: Simultaneous Geometry and Texture Editing for Point-Based Representations (ECCV 2026). Code: https://github.com/zvict/pointgt
Fetch with the release's own script, which verifies every file against a pinned sha256:
python download.py --what checkpoints # the seven NeRF-Synthetic scenes
python download.py --what demo # the dress editing demo
python download.py --verify # re-check what is already on disk
| Path | What it is |
|---|---|
checkpoints/nerfsyn/<scene>/model.pth | evaluated checkpoint for each of the seven scenes |
checkpoints/nerfsyn/<scene>/base.pth | pre-finetune parent for lego, materials, mic, ship |
checkpoints/nerfsyn/init/<scene>.ply | fixed initial point cloud where a config loads one |
demo/dress/papr.pth | stage-a point renderer for the dress scene |
demo/dress/nuvo_model_with_texture.ckpt | stage-b UV atlas and texture map |
demo/dress/texture_map_grid.png | learned texture, packed as a four-chart grid |
demo/dress/rbf_pcds/edit_*.ply | 35-frame deformed point-cloud sequence |
demo/dress/control_points.ply | control points the stage-a run was initialised from |
data/sketchfab/dress1/ | rendered views and cameras for the dress scene |
The dress scene is derived from Dress with gold leaves by
Canvastique3D, licensed
CC BY 4.0. The renders, point clouds, checkpoints and
texture under demo/dress/ and data/sketchfab/dress1/ are derived works; keep that credit if you
use them.
NeRF Synthetic is not redistributed here. Get it from the original NeRF release.
Code in the linked repository is MIT.