# niessner/Matterport

Matterport3D is a pretty awesome dataset for RGB-D machine learning tasks :)

Repository: https://github.com/niessner/Matterport
Canonical: https://ross.abutalabs.com/products/matterport
Homepage: https://niessner.github.io/Matterport/
Language: C++
License: MIT
License Family: permissive
Topics: 3d-reconstruction, semantic-scene-understanding
Last push: 2025-11-03T12:14:52+00:00

## Health v2 (maintenance only)
Score: 55/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 50, release rhythm 35, longevity 100
- inputs: {"age_days": 3548, "days_push": 303, "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 1245, forks 156 (observed 2026-08-28T04:04:06.955459+00:00)

## What it is
Matterport3D is a large-scale RGB-D dataset of 90 building-scale indoor scenes with 194,400 RGB-D images, panoramic views, textured 3D meshes, camera poses, and 2D/3D semantic annotations. The repository also provides code, loaders, and benchmark tasks for scene understanding research.

## Use cases
- train RGB-D scene understanding models
- semantic segmentation of indoor 3D scenes
- keypoint matching and view overlap prediction benchmarks
- surface normal estimation from color images
- 3D reconstruction research with aligned RGB-D data
- region type classification and semantic voxel labeling

## When to choose
- you need large-scale indoor RGB-D data with 3D semantic annotations
- you are benchmarking scene understanding or 3D vision algorithms
- you need globally aligned panoramic views of entire buildings

## When to avoid
- you need outdoor or non-building-scale scenes
- you cannot sign the institutional Terms of Use agreement required for data access
- you need a lightweight dataset for quick prototyping

## Facets
- artifact type: dataset
- maturity: stable
- function: computer-vision, image-processing, machine-learning, data-science
- domain: computer-vision, machine-learning, deep-learning, artificial-intelligence
- platform: python, cpp, cross-platform
- tags: rgb-d, 3d-reconstruction, semantic-segmentation, indoor-scenes, benchmark, point-clouds, meshes, scene-understanding

## Member repositories
- niessner/Matterport (main) score 55

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:06.955459+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 2026-08-30T05:08:10.925688+00:00, confidence not recorded.
  - readme: https://github.com/niessner/Matterport (fetched 2026-08-28T04:04:06.955459+00:00, sha b3e05b5108c6)
  - homepage: https://niessner.github.io/Matterport/ (fetched 2026-08-29T12:19:34.291647+00:00, sha dc92e272c54f)
- Data as of 2026-08-30T08:39:29.467469+00:00.