CognitiveDrone: A VLA Model and Evaluation Benchmark for Real-Time Cognitive Task Solving and Reasoning in UAVs

Abstract

This paper introduces CognitiveDrone, a novel Vision-Language-Action (VLA) model tailored for complex Unmanned Aerial Vehicles (UAVs) tasks that demand advanced cognitive abilities. Trained on a dataset comprising over 8,000 simulated flight trajectories across three key categories—Human Recognition, Symbol Understanding, and Reasoning—the model generates real-time 4D action commands based on first-person visual inputs and textual instructions. To further enhance performance in intricate scenarios, we propose CognitiveDrone-R1, which integrates an additional Vision-Language Model (VLM) reasoning module to simplify task directives prior to high-frequency control. Experimental evaluations using our open-source benchmark, CognitiveDroneBench, reveal that while a racing-oriented model (RaceVLA) achieves an overall success rate of 31.3%, the base CognitiveDrone model reaches 59.6%, and CognitiveDrone-R1 attains a success rate of 77.2%. These results demonstrate improvements of up to 30% in critical cognitive tasks, underscoring the effectiveness of incorporating advanced reasoning capabilities into UAV control systems. Our contributions include the development of a state-of-the-art VLA model for UAV control and the introduction of the first dedicated benchmark for assessing cognitive tasks in drone operations. The complete repository is available at CognitiveDrone.

Dataset Structure

data/rlds/ – Data for training and validation in RLDS format:
- train/ – Training data.
data/benchmark/ – Data for the simulation benchmark:
- validation – JSON files for model evaluation.

Instructions for Use

Training: Use the data from data/rlds/train/ for model training.
Evaluation: Run the simulation benchmark using the files from data/benchmark/validation/.

Links and Bibliography

Project Repository: CognitiveDrone
Paper: BibTeX reference will be available soon (coming soon).

This provides you with a complete package of files that can be easily uploaded to Hugging Face. Once the repository is created on the platform, you can use the git lfs command for large files if necessary.

If you have any further questions regarding repository organization or the dataset card, feel free to ask!