VL3D++ is an open framework, documented and reproducible
All the code, the architecture and the examples
are publicly available.
What the framework is
VL3D++ is a software framework for point-by-point classification and regression on 3D point clouds.
It allows running the full workflow on 3D data:
Model training
Prediction on new clouds
Automatic data preparation
Geometric feature extraction
Model and result evaluation
It also generates automatically:
Charts
Text reports with quantitative results.
Point clouds with viewable results
Architecture and operation
It is designed so that the user can configure and run experiments without modifying the code, through JSON files that define models, parameters and evaluations.
The framework is organised to cover the entire pipeline:
VL3D++ implements different neural network architectures for 3D point clouds published in high-impact scientific journals:
KPConv (Kernel Point Convolution)
Deformable and flexible convolution operator on points.
Sparse 3D Convolutional Classifier
Efficient implementation through sparse convolutions.
SFL-NET (Slight Filter Learning Network)
Lightweight network optimised for large-scale segmentation.
Point Transformer (V1/V2)
Models based on attention mechanisms (Self-attention).
Origin of the framework
VL3D++ is an evolution of the original VirtuaLearn3D (VL3D) framework, developed by the 3DGeo research group in a project funded by the German Research Foundation (DFG).
Repository and documentation
GitLab
Access to the full source code under the MIT licence.
VL3D++ Repository
Documentation
Technical guides, detailed tutorials.
Official Documentation
Reproducible examples
Pre-configured pipelines ready to train and predict on various datasets. Examples
Licence
The VL3D++ framework is distributed under the MIT licence.
The documentation of VirtuaLearn3D++ is published under the CC BY 4.0 licence by Alberto Manuel Esmorís Pena, allowing its use and study with attribution.
Do you have a point cloud project?
Collaboration, research and technology transfer
We foster an ecosystem of open innovation and continuous collaboration between academia and industry. Our technical and research team can guide you to explore how to apply VL3D++ to your case, collaborate on research or study technology-transfer pathways. Anyone interested in professional point cloud classification, or who needs the development of a specific technical functionality for their workflows, can contact our development team directly. We are open to joint research projects, technology transfer and advanced support for the implementation of the framework in production environments.
