Fetching the paper…

LED: LLM Enhanced Open-Vocabulary Object Detection without Human Curated Data Generation · Around