Extreme Driving Conditions Dataset with Multi-Modal Sensor Data from Tsinghua University

A multi-modal driving dataset focused on extreme, critical, and adverse-condition driving scenarios. It was released by the Intelligent Chassis Team at Tsinghua University's School of Vehicle and Mobility. The dataset includes human driving data as well as autonomous driving data, last updated on the platform in April 2026.

Use Cases

• Training perception models for adverse weather based on synchronized camera and LiDAR data.
• Benchmarking autonomous driving algorithms in critical scenarios using vehicle dynamics and GPS state.
• Developing sensor fusion techniques for extreme conditions based on the multi-modal sensor rig.

Strengths

• Data collection uses a multi-sensor rig including surround cameras, a stereo camera with depth, a forward-facing LiDAR, and synchronized vehicle dynamics.
• Focus on extreme, critical, and adverse-condition driving scenarios, which are less common in public datasets.
• Includes both human driving data and autonomous driving data for comparative analysis.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Intelligent Chassis Team, School of Vehicle and Mobility, Tsinghua University.

Collection Method

Collected and curated with a surround camera rig, a ZED 2i stereo camera (with depth), a forward-facing LiDAR, and synchronized vehicle dynamics / GPS state.

Time Range

null

Freshness

Last updated 2026-04-27 08:33:52; freshness should be verified.

Geography

null