Update README.md

README.md

@@ -17,6 +17,17 @@ This model is a **Multi-Task Image Classifier** that performs two tasks simultan
 
 The model uses a **ResNet-50** backbone with two heads: one for multi-class object recognition and another for binary classification (AI-generated vs. Real). It was trained on a subset of the [Hemg/AI-Generated-vs-Real-Images-Datasets](https://huggingface.co/datasets/Hemg/AI-Generated-vs-Real-Images-Datasets) and leverages YOLO for improved pseudo-labeling across the entire dataset.
 
+## Model Details
+- **Trained by:** [Abdellahi El Moustapha](https://abmstpha.github.io/)
+- **Programming Language:** Python
+- **Base Model:** ResNet-50
+- **Datasets:** Hemg/AI-Generated-vs-Real-Images-Datasets
+- **Library:** PyTorch
+- **Pipeline Tag:** image-classification
+- **Metrics:** Accuracy for both binary classification and multi-class object recognition
+- **Version:** v1.0
+
+
 ## Intended Use
 This model is designed for:
 - **Digital Content Verification:** Detecting AI-generated images to help prevent misinformation.
@@ -47,15 +58,7 @@ Alternatively, you can test the model directly via our interactive demo:
   - **Object Recognition:** Pseudo-label accuracy started at around 38–40% and improved during training.
 - **Evaluation:** Detailed evaluation metrics and loss curves are available in our training logs.
 
-## Model Details
-- **Trained by:** [Abdellahi El Moustapha](https://abmstpha.github.io/)
-- **Programming Language:** Python
-- **Base Model:** ResNet-50
-- **Datasets:** Hemg/AI-Generated-vs-Real-Images-Datasets
-- **Library:** PyTorch
-- **Pipeline Tag:** image-classification
-- **Metrics:** Accuracy for both binary classification and multi-class object recognition
-- **Version:** v1.0
+
 
 ## Limitations and Ethical Considerations
 - **Pseudo-Labeling:** The object recognition task uses pseudo-labels generated from a pretrained model, which may introduce noise or bias.