OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

3D Bonet

3D Bonet is an MIT-licensed Python research codebase that uses TensorFlow to learn object bounding boxes for instance segmentation on 3D point clouds. It includes training and evaluation scripts and documents results and visualization workflows for ScanNet and S3DIS.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

3D Bonet — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it as a research codebase when you can work with its documented legacy Python, TensorFlow and CUDA environment and have suitable point-cloud data.

Who it is for

  • Researchers studying 3D point-cloud instance segmentation.
  • Developers building experiments around object bounding-box prediction for indoor-scene point clouds.

Documented uses

  • Train a point-cloud instance-segmentation model for indoor scenes.
  • Generate object bounding-box predictions from 3D point-cloud data.
  • Evaluate results using the documented ScanNet or S3DIS workflows.
Before you choose

Requirements & limits

The public GitHub repository specifies an MIT license.

  • The README specifies Ubuntu 16.04, CUDA 8.0 and particular older Python and TensorFlow versions.
  • TensorFlow point-cloud operations require path adjustments and compilation.
  • Training and evaluation are provided as repository scripts rather than as a hosted service.

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

Prepare the documented environment and compile TensorFlow operations

Set up the listed dependencies and compile the repository's sampling operation before using its training or evaluation scripts.

Before you start

  • Ubuntu 16.04 and CUDA 8.0
  • Python 2.7 or 3.6
  • TensorFlow 1.2 or 1.4, SciPy 1.3, h5py 2.9 and Open3D-Python 0.3.0

Steps

  1. Install the documented system and Python dependencies.

  2. Use TensorFlow's sysconfig functions to locate its include and library paths.

  3. Update paths in the TensorFlow operation compilation files, including tf_ops/sampling/tf_sampling_compile.sh.

  4. From tf_ops/sampling, make tf_sampling_compile.sh executable and run it.

Expected outcome

The documented sampling TensorFlow operation is compiled as a setup stage for the codebase.

Guide limitations

  • The README documents this setup but does not establish compatibility with newer dependency versions.
  • This guide covers environment preparation and compilation, not a verified training or evaluation run.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
Yang7879/3D-BoNet · Official homepage title · retained in its source language
How we review sourcesSuggest a correction