OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Torch Points3D

Torch Points3D is a Python framework built around PyTorch for developing, training and applying deep-learning models to 3D point-cloud data. It organizes models and datasets by task and includes configurations, metrics, visualization tools and inference examples.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Torch Points3D — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Torch Points3D when you need a configurable PyTorch workflow for 3D point-cloud models, from task-based training to inference on custom point-cloud data.

Who it is for

  • Researchers and developers building point-cloud deep-learning experiments.
  • Teams working on tasks such as segmentation, classification, registration or 3D object detection.

Documented uses

  • Configure and train point-cloud models for a selected task.
  • Explore methods including PointNet, PointNet++, KPConv and VoteNet.
  • Apply a pretrained model to custom, potentially unannotated point clouds.
  • Inspect inference results using the provided visualization notebook.
Before you choose

Requirements & limits

The project is distributed as a Python package and its repository documents installation with pip.

  • The framework relies heavily on PyTorch Geometric and Hydra.
  • GPU use requires CUDA 10 or later; sparse-convolution support uses an optional backend with separate setup instructions.
  • The README recommends Docker to ease setup when working with specific CUDA and PyTorch versions.
  • The repository includes a training script, an inference script, task configurations, and a Dockerfile and build script for containerizing trained models.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up Torch Points3D

Install the Python package after preparing a supported Python and PyTorch environment.

Before you start

  • Python 3.7 or later with Python headers.
  • PyTorch 1.8.1 or later; PyTorch 1.9 or later is recommended.
  • CUDA 10 or later if using the GPU version.

Steps

  1. Prepare the environment natively or with Docker; the README recommends Docker for easing setup with specific CUDA and PyTorch versions.

  2. Install the package with the command below.

    Documented command
    pip install torch-points3d

Expected outcome

The Torch Points3D Python package is installed in the prepared environment.

Guide limitations

  • A sparse-convolution backend is optional and has separate installation instructions.
  • This setup procedure does not itself train a model or run inference.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
torch-points3d/torch-points3d · Official homepage title · retained in its source language
How we review sourcesSuggest a correction