🚀 Mirage-Photo-Classifier
Mirage-Photo-Classifier is an image classification vision - language encoder model fine - tuned from google/siglip2 - base - patch16 - 224 for binary image authenticity classification. It can determine if an image is real or AI - generated.

✨ Features
- Binary Classification: Mirage-Photo-Classifier is an image classification vision - language encoder model fine - tuned from google/siglip2 - base - patch16 - 224 for a binary image authenticity classification task. It uses the SiglipForImageClassification architecture to determine whether an image is real or AI - generated (fake).
- High Performance: The model shows high precision and recall in the classification report.
Classification Report:
precision recall f1-score support
Real 0.9781 0.9132 0.9446 5000
Fake 0.9186 0.9796 0.9481 5000
accuracy 0.9464 10000
macro avg 0.9484 0.9464 0.9463 10000
weighted avg 0.9484 0.9464 0.9463 10000
- Clear Categorization: The model categorizes images into two classes:
- Class 0: Real
- Class 1: Fake

📦 Installation
!pip install -q transformers torch pillow gradio
💻 Usage Examples
Basic Usage
import gradio as gr
from transformers import AutoImageProcessor
from transformers import SiglipForImageClassification
from PIL import Image
import torch
model_name = "prithivMLmods/Mirage-Photo-Classifier"
model = SiglipForImageClassification.from_pretrained(model_name)
processor = AutoImageProcessor.from_pretrained(model_name)
labels = {
"0": "Real",
"1": "Fake"
}
def classify_image_authenticity(image):
"""Predicts whether the image is real or AI-generated (fake)."""
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()
predictions = {labels[str(i)]: round(probs[i], 3) for i in range(len(probs))}
return predictions
iface = gr.Interface(
fn=classify_image_authenticity,
inputs=gr.Image(type="numpy"),
outputs=gr.Label(label="Prediction Scores"),
title="Mirage Photo Classifier",
description="Upload an image to determine if it's Real or AI-generated (Fake)."
)
if __name__ == "__main__":
iface.launch()
📚 Documentation
Intended Use
The Mirage - Photo - Classifier model is designed to detect whether an image is genuine (photograph) or synthetically generated. Use cases include:
- AI Image Detection: Identifying AI - generated images in social media, news, or datasets.
- Digital Forensics: Helping professionals detect image authenticity in investigations.
- Platform Moderation: Assisting content platforms in labeling generated content.
- Dataset Validation: Cleaning and verifying training data for other AI models.
📄 License
The model is licensed under the Apache - 2.0 license.
Property |
Details |
Model Type |
Image classification vision - language encoder model |
Training Data |
anson - huang/mirage - news |
Base Model |
google/siglip2 - base - patch16 - 224 |
Pipeline Tag |
image - classification |
Library Name |
transformers |
Tags |
Fake, Real, SigLIP2, Mirage |