VVSim: Large-Scale Aerial-Ground Cooperative Perception Dataset

61,000 fully annotated frames collected for aerial-ground cooperative perception. The dataset integrates synchronized multimodal sensing data and state information from vehicles and UAVs, covering 19 interaction scenarios and 5 weather conditions. It was created by LOTEAT and last updated on Hugging Face in April 2026.

Use Cases

• Training vehicle detection models based on synchronized multimodal sensor data.
• Developing trajectory prediction algorithms based on annotated interaction scenarios like cut-in and lane change.
• Testing perception system robustness based on data covering 5 weather conditions such as foggy and snowy.
• Researching aerial-ground cooperative perception (AGCP) based on integrated vehicle and UAV state information.
• Semantic segmentation of driving scenes based on annotated frames across 11 scene types like city and highway.

Strengths

• Contains 61,000 fully annotated frames, indicating a substantial scale.
• Covers 19 distinct interaction scenarios, providing variety in driving behaviors.
• Includes data from 5 different weather conditions, which may help test model robustness.
• Integrates synchronized data from both vehicles and UAVs for cooperative perception.

Limitations

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

Provenance

Source

LOTEAT via Hugging Face.

Collection Method

Integrated synchronized multimodal sensing data and state information collected from vehicles and UAVs.

Time Range

null

Freshness

Last updated 2026-04-03 08:35:15; freshness should be verified.

Geography

Scene types include city, town, university, highway, and mountain, but specific geographic coverage is not stated.