Descriptive Caption Enhancement with Visual Specialists for Multimodal Perception

If you like our project, please give us a star ⭐ on GitHub.
Please download the required image files directly from the [URL_images](https://huggingface.co/datasets/BAAI/DenseFusion-1M/tree/main/images/DenseFusion-1M) and store them in a folder named `DCE-1M-images` as specified. Ensure that the folder path is correct to enable subsequent operations to proceed smoothly.