# ranahanocka/point2mesh

Reconstruct Watertight Meshes from Point Clouds [SIGGRAPH 2020]

Repository: https://github.com/ranahanocka/point2mesh
Canonical: https://ross.abutalabs.com/products/point2mesh
Language: Python
License: MIT
License Family: permissive
Topics: 3d-graphics, deep-learning, mesh-processing, pointcloud, pytorch, reconstruction
Last push: 2023-11-03T12:41:58+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2308, "days_push": 1034, "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 1240, forks 135 (observed 2026-08-28T04:04:06.166355+00:00)

## What it is
Point2Mesh is a PyTorch implementation of a SIGGRAPH 2020 technique that reconstructs watertight surface meshes from input point clouds by optimizing a CNN to deform an initial mesh around the shape. It is research code accompanying an academic paper, with example scripts and conda-based setup.

## Use cases
- reconstruct a watertight mesh from a 3d point cloud
- shrink-wrap an initial mesh onto point cloud data using a learned cnn
- reproduce the point2mesh siggraph 2020 paper results
- process 3d scans into clean manifold meshes
- experiment with deep learning for geometry processing

## When to choose
- you need watertight mesh reconstruction from point clouds of a single object
- you want to study or extend the point2mesh self-prior approach
- you have a GPU and can work with PyTorch 1.4/1.5 and PyTorch3D

## When to avoid
- you need a fast, production-ready reconstruction pipeline
- you cannot install the pinned old PyTorch/PyTorch3D versions or the external Manifold tool
- you need generic multi-object or real-time reconstruction

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning, graphics, image-processing
- domain: computer-vision, graphics, deep-learning, machine-learning
- platform: python
- tags: point-cloud, mesh-reconstruction, pytorch, 3d-graphics, siggraph, watertight-mesh, research-code, linux, gpu

## Member repositories
- ranahanocka/point2mesh (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:06.166355+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 2026-08-30T05:09:54.392796+00:00, confidence not recorded.
  - readme: https://github.com/ranahanocka/point2mesh (fetched 2026-08-28T04:04:06.166355+00:00, sha 74e70befd4d4)
- Data as of 2026-08-30T08:39:29.467469+00:00.