File size: 1,444 Bytes
5d677cb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8c9d36b
5d677cb
8c9d36b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
---
tags:
- vision
- image-classification
- vit
- transformer
- fake-image-detection
license: apache-2.0
datasets:
- ciplab/real-and-fake-face-detection
model-index:
- name: SahAi
  results:
  - task:
      name: Image Classification
      type: image-classification
    dataset:
      name: Real and Fake Face Detection Dataset
      type: ciplab/real-and-fake-face-detection
    metrics:
      - name: Accuracy
        type: accuracy
        value: 99.12
      - name: Precision
        type: precision
        value: 98.95
      - name: Recall
        type: recall
        value: 99.00
---

# SahAi - Enhanced Fake Image Detection Model
**SahAi** is a fine-tuned Vision Transformer (ViT) model designed for fake image localization in social media.

### 🚀 Model Description
- **Base Model:** ViT-B/16  
- **Input Size:** 224x224  
- **Output Classes:** Real (0), Fake (1)  

### 🔥 How to Use
```python
from transformers import ViTForImageClassification, AutoFeatureExtractor
from PIL import Image
import torch

model = ViTForImageClassification.from_pretrained("SahilSha/SahAi")
feature_extractor = AutoFeatureExtractor.from_pretrained("google/vit-base-patch16-224-in21k")

image = Image.open("test_image.jpg").convert("RGB")
inputs = feature_extractor(images=image, return_tensors="pt")

with torch.no_grad():
    outputs = model(**inputs)
    prediction = outputs.logits.argmax(-1).item()

print("Real" if prediction == 0 else "Fake")