The Driving Scene Knowledge Graph Constructor (DSKG Constructor) is an Open Source Project that support the automated creation of RDF Knowledge Graphs (KG) from publicly available autonomous driving datasets (e.g. nuScenes, Pandaset, BDD).
The DSKG Constructor aims to support the research community to create RDF Knowledge
Graphs from publicly available autonomous driving dataset for the purpose of
developing ML/DL architectures and learning models on the knowledge graph for
various autonomous driving task, e.g. object prediction, motion prediction.
The software is not ready for production use. It has neither been developed nor tested for a specific use case. However, the license conditions of the applicable Open Source licenses allow you to adapt the software to your needs. Before using it in a safety relevant setting, make sure that the software fulfills your requirements and adjust it according to any applicable safety standards (e.g. ISO 26262).
The DSKG Constructor aims to support the creation of a Driving Scene KG from various publicly available autonomous driving datasets.
For each supported dataset type, we include all the required python scripts along with the respective ontology for that dataset and the documentation (README.md) that instructs how to run the script(s) in the respective sub-folder.
Please note that this project provides only the software to construct the knowledge graphs from the various source datasets. Due to the various licensing conditions of the source datasets, we just provide you with the download links in the respective README file.
The permission to use the dataset is NOT covered by the license of the DSKG Constructor. You MUST adhere to the license conditions set by the providers of the source datasets.
The DSKG Constructor software is provided under the Apache-2.0 license. See the LICENSE file for details.
For a list of other open source components included, see the file 3rd-party-licenses.txt.