The dataset is comprised of a 6, 051 unique UAV configurations, where multiple data formats, including a grammar string, an RGB image, and an GLB file describe each configuration.
Complementing these representation modalities, we also provide a configuration-based description, i.e., a text descriptor describing the features of each UAV using natural language.
