🛡️ TruthLens AI

AI-Powered Deepfake Detection for Images & Videos

TruthLens AI is a full-stack web application that detects whether an image or video is REAL or FAKE using advanced AI/ML models and forensic techniques like Error Level Analysis (ELA).

---

DEMO - {https://youtu.be/7M6BlRiKiGc?si=HYMEWQCKeC0IeNVi}

---

🚀 Features

• Deepfake Detection (Images & Videos)
• AI Model Integration (Hugging Face Transformer)
• ELA (Error Level Analysis) Heatmap
• Confidence Score + Risk Level
• Frame-based Video Analysis
• Google Sign-In (optional)
• Downloadable Analysis Report
• Modern UI (Next.js)

---

🧠 AI Model Used

Hugging Face Model:
https://huggingface.co/dima806/deepfake_vs_real_image_detection

• Type: Image Classification Transformer
• Task: Detect REAL vs FAKE images
• Detects:
  • Compression artifacts
  • GAN inconsistencies
  • Pixel-level anomalies

---

🏗️ Tech Stack

Frontend:

• Next.js (React)
• JavaScript / TypeScript
• CSS

Backend:

• Flask (Python)
• OpenCV
• Pillow
• Transformers (Hugging Face)
• PyTorch

---

📁 Project Structure

TruthLens-AI/

front/ → Next.js frontend
back/ → Flask backend

app.py                → Main backend server  
model/detector.py     → AI model logic  
utils/ela.py          → ELA generation  
utils/video.py        → Video analysis  

---

⚙️ Setup Instructions

1. Clone Repository

git clone
cd TruthLens-AI

---

2. Backend Setup

cd back

Create virtual environment: python3 -m venv venv
source venv/bin/activate

Install dependencies: pip install flask flask-cors transformers torch torchvision torchaudio opencv-python pillow

Run backend: python app.py

Backend runs on: http://127.0.0.1:5000

---

3. Frontend Setup

cd front

Install dependencies: npm install

Run frontend: npm run dev

Frontend runs on: http://localhost:3000

---

🔗 API Endpoints

POST /analyze-image

Input: file

Output: { "type": "image", "label": "FAKE", "confidence": 87.2, "original_image": "path", "ela_image": "path" }

---

POST /analyze-video

Output: { "type": "video", "label": "REAL", "confidence": 72.5, "video_path": "path" }

---

🔬 How It Works

1. User uploads image or video
2. Frontend sends file to Flask backend
3. Backend:
   • Saves file
   • Runs AI model
   • Generates ELA (for images)
4. Returns prediction + confidence
5. Frontend displays result and visualization

---

⚠️ Important Notes

• Use Python 3.10 (not 3.13)
• First run will download model
• Works on CPU
• Video analysis uses sampled frames

---

📸 Demo Flow

1. Open website
2. Navigate pages
3. Login with Google
4. Upload file
5. View result + confidence
6. Toggle ELA heatmap
7. Download report

---

📌 Future Improvements

• Real-time webcam detection
• Cloud deployment
• Mobile optimization
• Better video models
• Advanced reports

---