OpenUAV-QA is a large-scale multiple-choice question-answering benchmark for UAV (drone) navigation decision-making, built upon the TravelUAV (OpenUAV) dataset. It transforms raw UAV flight trajectories into structured, text-polished 4-option QA pairs that test a multimodal model's ability to reason about path planning, action sequences, and spatial dynamics from first-person drone video footage. The dataset was created by choucsan and was last updated on 2026-07-12.
Use Cases
- Benchmarking multimodal reasoning for UAV path planning based on flight trajectory data.
- Training models to understand spatial dynamics from first-person drone video footage.
- Evaluating AI on structured, text-polished multiple-choice questions about action sequences.
Strengths
- Large-scale benchmark for UAV navigation decision-making.
- Contains structured, text-polished 4-option QA pairs.
- Built upon the TravelUAV (OpenUAV) dataset of flight trajectories.
Limitations
- Row count is unknown, which may limit suitability assessment.
- Column-level documentation is absent; field semantics must be inferred after download.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- choucsan on Hugging Face.
- Collection Method
- Transformed from the TravelUAV (OpenUAV) dataset of raw UAV flight trajectories.
- Freshness
- Last updated 2026-07-12 14:28:21; freshness should be verified.