Advancing Grounded Multimodal Named Entity Recognition via LLM-Based Reformulation and Box-Based Segmentation 2406.07268 2024-06-11 이 페이지의 평가 1건 F1: 63.92