Update README.md

Former-commit-id: 11dc173776f538e2510f50b202755e763f8267cb

README.md

@@ -74,6 +74,27 @@ After that, input the text prompt and then the image path. For example，
 The results should be like:
 <p align="center"> <img src="imgs/example1.jpg" width="22%"> <img src="vis_output/example1_masked_img_0.jpg" width="22%"> <img src="imgs/example2.jpg" width="25%"> <img src="vis_output/example2_masked_img_0.jpg" width="25%"> </p>
 
+## Dataset
+We have collected 1218 images, i.e., 239 train, 200 val, and 779 test. The training and validation sets can be download from <a href="https://drive.google.com/drive/folders/125mewyg5Ao6tZ3ZdJ-1-E3n04LGVELqy?usp=sharing">**this link**</a>. 
+
+Each image is provided with an annotation JSON file:
+```
+image_1.jpg, image_1.json
+image_2.jpg, image_2.json
+...
+image_n.jpg, image_n.json
+```
+Important keys contained in JSON files:
+```
+- "text": text instructions.
+- "is_sentence": whether the text instructions are long sentences.
+- "shapes": target polygons.
+```
+
+The elements of the "shapes" exhibit two categories, namely **"target"** and **"ignore"**. The former category is indispensable for evaluation, while the latter category denotes the ambiguous region and hence disregarded during the evaluation process. 
+
+Besides, we leveraged GPT-3.5 for rephrasing instructions, so images in the training set may have **more than one instructions (but fewer than six)** in the "text" field. Users can randomly select one instruction as the text query to obtain a better model.
+
 
 ## Citation 
 If you find this project useful in your research, please consider citing: