The rapid advancement of large language models (LLMs) and multimodal learning has
transformed digital content creation and manipulation. Traditional visual editing tools require …