README.md

---
license: apache-2.0
language:
- en
base_model:
- google/siglip2-base-patch16-224
pipeline_tag: image-classification
library_name: transformers
tags:
- deepfake
- detection
---
#  **Deepfake-Detect-Siglip2**

**Deepfake-Detect-Siglip2** is an image classification vision-language encoder model fine-tuned from google/siglip2-base-patch16-224 for a single-label classification task. It is designed to detect whether an image is real or a deepfake using the SiglipForImageClassification architecture.  

The model categorizes images into two classes:  
- **Class 0:** "Fake" – The image is detected as a deepfake or manipulated.  
- **Class 1:** "Real" – The image is classified as authentic and unaltered.  

# **Run with Transformers🤗**

```python
!pip install -q transformers torch pillow gradio
```

```python
import gradio as gr
from transformers import AutoImageProcessor
from transformers import SiglipForImageClassification
from transformers.image_utils import load_image
from PIL import Image
import torch

# Load model and processor
model_name = "prithivMLmods/Deepfake-Detect-Siglip2"
model = SiglipForImageClassification.from_pretrained(model_name)
processor = AutoImageProcessor.from_pretrained(model_name)

def deepfake_detection(image):
    """Classifies an image as Fake or Real."""
    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()
    
    labels = model.config.id2label
    predictions = {labels[i]: round(probs[i], 3) for i in range(len(probs))}
    
    return predictions

# Create Gradio interface
iface = gr.Interface(
    fn=deepfake_detection,
    inputs=gr.Image(type="numpy"),
    outputs=gr.Label(label="Detection Result"),
    title="Deepfake Detection Model",
    description="Upload an image to determine if it is Fake or Real."
)

# Launch the app
if __name__ == "__main__":
    iface.launch()
```

# **Intended Use:**  

The **Deepfake-Detect-Siglip2** model is designed to distinguish between **real and fake (deepfake) images**. It is useful for identifying AI-generated or manipulated content.  

### Potential Use Cases:  
- **Deepfake Detection:** Identifying AI-generated fake images.  
- **Content Verification:** Assisting social media platforms in filtering manipulated content.  
- **Forensic Analysis:** Supporting cybersecurity and investigative research on fake media.  
- **Media Authenticity Checks:** Helping journalists and fact-checkers verify image credibility.