Semantic Scene Understanding with Large Language Models on Unmanned Aerial Vehicles

Unmanned Aerial Vehicles (UAVs) are able to provide instantaneous visual cues and a high-level data throughput that could be further leveraged address complex tasks, such as semantically rich scene understanding. In this work, we built on the use of Large Language Models (LLMs) Visual (VLMs), together with state-of-the-art detection pipeline, thorough zero-shot UAV literary text descriptions. The generated texts achieve GUNNING Fog median grade level in range 7–12. Applications framework found filming industry enhance user experience theme parks or advertisement sector. We demonstrate low-cost highly efficient practical implementation microdrones well-controlled challenging setting, addition proposing standardized readability metrics assess LLM-enhanced