# drprojects/superpoint_transformer

Official PyTorch implementation of Superpoint Transformer [ICCV'23], SuperCluster [3DV'24 Oral], and EZ-SP [ICRA'26]

Repository: https://github.com/drprojects/superpoint_transformer
Canonical: https://ross.abutalabs.com/products/superpoint_transformer
Language: Python
License: MIT
License Family: permissive
Topics: 3d, deep-learning, efficient, fast, hierarchical, lightweight, partition, point-cloud, pytorch, semantic-segmentation, superpoint, transformer, graph-clustering, panoptic-segmentation, partitioning, 3dv2024, iccv2023
Last push: 2026-04-21T16:42:35+00:00

## Health v2 (maintenance only)
Score: 64/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 78, release rhythm 35, longevity 83
- inputs: {"age_days": 1175, "days_push": 134, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1049, forks 133 (observed 2026-08-28T04:03:22.629594+00:00)

## What it is
Official PyTorch implementation of Superpoint Transformer (ICCV'23), SuperCluster (3DV'24), and EZ-SP (ICRA'26) for efficient semantic and panoptic segmentation of large-scale 3D point clouds. It partitions point clouds into hierarchical superpoints and applies lightweight transformer-based segmentation on top.

## Use cases
- segment large 3d point clouds semantically
- run panoptic segmentation on lidar scans
- partition point clouds into superpoints
- train a fast lightweight 3d segmentation model
- cluster superpoints into object instances
- reproduce iccv 2023 superpoint transformer results

## When to choose
- you need efficient semantic or panoptic segmentation of large-scale 3D scenes
- you want a lightweight, fast transformer model for point clouds
- you need hierarchical superpoint partitioning of point clouds
- you want reproducible research code from published papers

## When to avoid
- you need 2D image segmentation rather than 3D point clouds
- you need a production-ready plug-and-play inference service rather than a research codebase
- you work outside PyTorch or lack GPU resources

## Facets
- artifact type: library
- maturity: active
- function: deep-learning, machine-learning, image-processing
- domain: computer-vision, deep-learning, machine-learning
- platform: python, windows
- tags: point-cloud, semantic-segmentation, panoptic-segmentation, superpoint, transformer, pytorch, 3d-scene-understanding, graph-clustering, research-code, gpu, linux, macos

## Member repositories
- drprojects/superpoint_transformer (main) score 64

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:22.629594+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 2026-08-30T07:00:28.878157+00:00, confidence not recorded.
  - readme: https://github.com/drprojects/superpoint_transformer (fetched 2026-08-28T04:03:22.629594+00:00, sha 661378adbca8)
- Data as of 2026-08-30T08:39:29.467469+00:00.