Upload README.md with huggingface_hub

README.md

@@ -51,26 +51,38 @@ task_categories:
 tags:
 - dichotomous-image-segmentation
 - dis
-pretty_name: DIS
+- salient-object-detection
+- background-removal
+pretty_name: DIS5K
 size_categories:
 - 1K<n<10K
 ---
-# DIS5K
 
-This dataset has been ported from [DIS](https://github.com/xuebinqin/DIS) more specifically taken from [drive_link](https://drive.google.com/file/d/1O1eIuXX1hlGsV7qx4eSkjH231q7G1by1/view?usp=sharing)
+# DIS5K (Dichotomous Image Segmentation 5K)
 
-🔗 dataset page: https://xuebinqin.github.io/dis/index.html
+High-accuracy dichotomous image segmentation dataset, mirrored to the `nobg` org for convenience. Each example is an RGB image paired with a high-quality binary segmentation label.
 
-🔗 github repo : https://github.com/xuebinqin/DIS 
+- `image`: RGB source image
+- `label`: binary segmentation ground-truth mask
+- `image_name`: original image file name
+- `label_name`: original label file name
 
+Splits: `DIS_TR` (train, 3000), `DIS_VD` (validation, 470), and the four test sets `DIS_TE1`–`DIS_TE4` (500 each, increasing in structural complexity).
 
-## Citation 
-Bibtex
-```
+## Source & credit
+
+Ported from **DIS** ([github.com/xuebinqin/DIS](https://github.com/xuebinqin/DIS)), specifically from this [drive link](https://drive.google.com/file/d/1O1eIuXX1hlGsV7qx4eSkjH231q7G1by1/view?usp=sharing).
+
+- 🔗 Dataset page: https://xuebinqin.github.io/dis/index.html
+- 🔗 GitHub repo: https://github.com/xuebinqin/DIS
+
+## Citation
+
+```bibtex
 @InProceedings{qin2022,
-      author={Xuebin Qin and Hang Dai and Xiaobin Hu and Deng-Ping Fan and Ling Shao and Luc Van Gool},
-      title={Highly Accurate Dichotomous Image Segmentation},
-      booktitle={ECCV},
-      year={2022}
+  author={Xuebin Qin and Hang Dai and Xiaobin Hu and Deng-Ping Fan and Ling Shao and Luc Van Gool},
+  title={Highly Accurate Dichotomous Image Segmentation},
+  booktitle={ECCV},
+  year={2022}
 }
 ```