DrivIng is a large-scale multimodal dataset for driving research, featuring full digital twin integration. It was authored by Dominik Rößle and published via Harvard Dataverse in January 2026.
Use Cases
- Train multimodal perception models using synchronized sensor data and digital twin environment states.
- Develop and validate autonomous driving algorithms within a fully simulated digital twin counterpart.
- Conduct research on sensor fusion by leveraging the dataset's multimodal driving recordings.
Strengths
- Described as a large-scale dataset, indicating substantial data volume.
- Features full digital twin integration, providing a simulated counterpart to real-world data.
- Published and maintained on the Harvard Dataverse platform, a reputable data repository.
Limitations
- Specific scale metrics like row count, file size, and column details are unavailable.
- The dataset's temporal and geographic coverage are not specified.
Provenance
- Source
- Harvard Dataverse
- Freshness
- Last updated on January 22, 2026.