# NVIDIA/MinkowskiEngine

Minkowski Engine is an auto-diff neural network library for high-dimensional sparse tensors

Repository: https://github.com/NVIDIA/MinkowskiEngine
Canonical: https://ross.abutalabs.com/products/minkowskiengine
Homepage: https://nvidia.github.io/MinkowskiEngine
Language: Python
License: NOASSERTION
License Family: other
Topics: neural-network, computer-vision, sparse-tensors, convolutional-neural-networks, semantic-segmentation, auto-differentiation, spatio-temporal-analysis, space-time, deep-learning, 3d-convolutional-network, 4d-convolutional-neural-network, high-dimensional-data, high-dimensional-inference, trilateral-filter, 3d-vision, sparse-convolution, pytorch, minkowski-engine, cuda, sparse-tensor-network
Last push: 2024-03-05T16:00:19+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2709, "days_push": 911, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2956, forks 484 (observed 2026-08-28T04:07:32.178064+00:00)

## What it is
Minkowski Engine is an auto-differentiation neural network library for high-dimensional sparse tensors, built on PyTorch with CUDA acceleration. It provides standard layers such as convolution, pooling, and normalization for sparse tensor networks used in 3D vision tasks.

## Use cases
- run semantic segmentation on 3D point clouds
- train a 3D CNN for shape classification on ModelNet40
- reconstruct or complete 3D geometry from sparse scans
- build sparse convolution networks for high-dimensional data
- perform 4D spatio-temporal deep learning on video or scans

## When to choose
- you need efficient sparse convolutions on 3D/4D data within PyTorch
- you are doing point cloud segmentation, detection, or reconstruction research
- you want CUDA-accelerated coordinate management for sparse tensors

## When to avoid
- you need dense tensor networks or standard 2D vision models
- you require active maintenance or support for the latest PyTorch/CUDA versions
- you want a pure-Python solution without GPU/CUDA build requirements

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning, computer-vision, image-processing
- domain: deep-learning, computer-vision, machine-learning, graphics
- platform: python
- tags: sparse-tensors, sparse-convolution, pytorch, 3d-vision, point-clouds, auto-differentiation, cuda, neural-networks, linux, gpu

## Member repositories
- NVIDIA/MinkowskiEngine (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:32.178064+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:32:13.137657+00:00, confidence not recorded.
  - readme: https://github.com/NVIDIA/MinkowskiEngine (fetched 2026-08-28T04:07:32.178064+00:00, sha 134f84eef8a7)
  - homepage: https://nvidia.github.io/MinkowskiEngine (fetched 2026-08-29T09:47:33.857848+00:00, sha 955d2f046580)
  - registry_pypi: https://pypi.org/pypi/minkowskiengine/json (fetched 2026-08-29T09:47:33.860668+00:00, sha 2a79cc6b1850)
- Data as of 2026-08-30T08:39:29.467469+00:00.