QUVA-Lab/escnn
Equivariant Steerable CNNs Library for Pytorch https://quva-lab.github.io/escnn/
This library helps machine learning researchers and practitioners build neural networks that are better at recognizing patterns in images and 3D data, especially when those patterns might be rotated or reflected. You provide standard 2D images, 3D scans, or scientific data like vector or scalar fields, and it outputs neural network models that are more efficient and accurate because they inherently understand these geometric transformations. It's designed for those developing advanced computer vision or scientific modeling applications.
508 stars. No commits in the last 6 months. Available on PyPI.
Use this if you are building deep learning models for tasks involving 2D or 3D data where rotations, translations, or reflections are common, and you want your models to generalize better with less training data.
Not ideal if your primary focus is on standard machine learning tasks that don't heavily rely on geometric symmetries, or if you prefer a simpler, less specialized deep learning framework.
Stars
508
Forks
61
Language
Python
License
—
Category
Last pushed
Oct 31, 2024
Commits (30d)
0
Dependencies
8
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/QUVA-Lab/escnn"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
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.