Update README.md

README.md

@@ -23,31 +23,9 @@ This dataset contains curated images with an oil canvas painting style, specific
    - The [BLIP-2 model](https://huggingface.co/blog/blip-2) was used to generate detailed image captions.
    - Captions include style, visual elements, and thematic descriptions to enhance model learning.
 
-## 📂 Dataset Structure
-The dataset is organized in a tabular format as shown below:
-
-| image             | text                                         |
-|-------------------|----------------------------------------------|
-| impressionism_3.jpg  | a woman sitting in the grass               |
-| impressionism_14.jpg | a woman sitting in front of a bird         |
-| impressionism_22.jpg | a house with trees and a tree              |
-| impressionism_25.jpg | a bridge over a river with a tower in the back... |
-| impressionism_40.jpg | flowers and a fan on a table               |
-
-The dataset is stored as image files along with a `.csv` or `.json` file containing the captions:
-```
-├── images/
-│   ├── impressionism_3.jpg
-│   ├── impressionism_14.jpg
-│   ├── impressionism_22.jpg
-│   └── ...
-└── captions.csv (or captions.json)
-```
-- `images/`: Directory containing all Impressionism style images.
-- `captions.csv`: A CSV file with columns `image` and `text`, linking each image to its caption.
 
 ## 🎯 Data Fields
-- **Image:** Oil canvas style image file (PNG/JPG).
+- **Image:** Oil canvas style image file (JPG).
 - **Caption:** Textual description generated by BLIP-2.
 
 ## 🔍 Data Collection Process