Update README.md

README.md

@@ -2,8 +2,23 @@
 license: apache-2.0
 datasets:
 - AadityaJain/Fromula_text_classification
+language:
+- en
+base_model:
+- google/siglip2-base-patch16-224
+pipeline_tag: image-classification
+library_name: transformers
+tags:
+- Formula-Text-Detection
+- SigLIP2
 ---
 
+![3.png](https://cdn-uploads.huggingface.co/production/uploads/65bb837dbfb878f46c77de4c/lg90wKzVcHjnTXs8_EGCR.png)
+
+# **Formula-Text-Detection**
+
+> **Formula-Text-Detection** is a vision-language encoder model fine-tuned from **google/siglip2-base-patch16-224** for **binary image classification**. It is built using the **SiglipForImageClassification** architecture to distinguish between **mathematical formulas** and **natural text** in document or image regions.
+
 ```py
 Classification Report:
               precision    recall  f1-score   support
@@ -17,3 +32,82 @@ weighted avg     0.9991    0.9991    0.9991     11832
 ```
 
 ![download.png](https://cdn-uploads.huggingface.co/production/uploads/65bb837dbfb878f46c77de4c/OdNUMSb_utc_RBWd3Gjfq.png)
+
+---
+
+## **Label Space: 2 Classes**
+
+The model classifies each input image into one of the following categories:
+
+```
+Class 0: "formula"
+Class 1: "text"
+```
+
+---
+
+## **Install Dependencies**
+
+```bash
+pip install -q transformers torch pillow gradio
+```
+
+---
+
+## **Inference Code**
+
+```python
+import gradio as gr
+from transformers import AutoImageProcessor, SiglipForImageClassification
+from PIL import Image
+import torch
+
+# Load model and processor
+model_name = "prithivMLmods/Formula-Text-Detection"  # Replace with your model path if different
+model = SiglipForImageClassification.from_pretrained(model_name)
+processor = AutoImageProcessor.from_pretrained(model_name)
+
+# Label mapping
+id2label = {
+    "0": "formula",
+    "1": "text"
+}
+
+def classify_formula_or_text(image):
+    image = Image.fromarray(image).convert("RGB")
+    inputs = processor(images=image, return_tensors="pt")
+
+    with torch.no_grad():
+        outputs = model(**inputs)
+        logits = outputs.logits
+        probs = torch.nn.functional.softmax(logits, dim=1).squeeze().tolist()
+
+    prediction = {
+        id2label[str(i)]: round(probs[i], 3) for i in range(len(probs))
+    }
+
+    return prediction
+
+# Gradio Interface
+iface = gr.Interface(
+    fn=classify_formula_or_text,
+    inputs=gr.Image(type="numpy"),
+    outputs=gr.Label(num_top_classes=2, label="Formula or Text"),
+    title="Formula-Text-Detection",
+    description="Upload an image region to classify whether it contains a mathematical formula or natural text."
+)
+
+if __name__ == "__main__":
+    iface.launch()
+```
+
+---
+
+## **Intended Use**
+
+**Formula-Text-Detection** can be used in:
+
+- **OCR Preprocessing** – Improve document OCR accuracy by separating formulas from text.
+- **Scientific Document Analysis** – Automatically detect mathematical content.
+- **Educational Platforms** – Classify and annotate scanned materials.
+- **Layout Understanding** – Help AI systems interpret mixed-content documents.