vevenom/pytorchgeonodes
PyTorchGeoNodes is a PyTorch module for differentiable shape programs / procedural models in forms of graphs. It can automatically translate Blender geometry node models into PyTorch code. Originally, it was designed to simplify the integration of procedural shape programs into machine learning pipelines for 3D scene understanding.
This project helps 3D artists, designers, and researchers in computer vision automatically reconstruct and understand 3D objects from scenes using Blender-designed shape programs. It takes 3D scene data (like from ScanNet) or synthetic scenes as input, and outputs optimized parameters for procedural 3D models. The end user is anyone working with 3D object modeling, scene understanding, or generating editable 3D content for simulations or digital assets.
Use this if you need to reverse-engineer 3D objects from scans or existing scenes into editable, parameterized models, rather than static meshes.
Not ideal if your primary goal is simple 3D model retrieval or if you are not familiar with Blender's Geometry Nodes for defining procedural shapes.
Stars
44
Forks
1
Language
Python
License
—
Category
Last pushed
Nov 23, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/vevenom/pytorchgeonodes"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
pyg-team/pytorch_geometric
Graph Neural Network Library for PyTorch
a-r-j/graphein
Protein Graph Library
raamana/graynet
Subject-wise networks from structural MRI, both vertex- and voxel-wise features (thickness, GM...
pykale/pykale
Knowledge-Aware machine LEarning (KALE): accessible machine learning from multiple sources for...
dmlc/dgl
Python package built to ease deep learning on graph, on top of existing DL frameworks.