Fetching the paper…

FireEdit: Fine-grained Instruction-based Image Editing via Region-aware Vision Language Model · Around