UAVMLLM is the first vision-language multi-modal large language model baseline, specifically tailored to low-altitude UAV scenarios.