@misc{indiciae8a79babf6ab0,
  title = {VisionDirector: Vision-Language Guided Closed-Loop Refinement for Generative Image Synthesis},
  author = {Meng Chu and Senqiao Yang and Haoxuan Che and Suiyun Zhang and Xichen Zhang and Shaozuo Yu and Haokun Gui and Zhefan Rao and Dandan Tu and Rui Liu and Jiaya Jia},
  year = {2026},
  url = {https://arxiv.org/abs/2512.19243},
  note = {Source identifier: 2512.19243}
}