Fetching the paper…

LLM-Optic: Unveiling the Capabilities of Large Language Models for Universal Visual Grounding · Around