README.md

metadata

title: DeepTrust
emoji: 🛡️
colorFrom: blue
colorTo: indigo
sdk: streamlit
sdk_version: 1.31.0
app_file: app.py
pinned: false
license: mit

🛡️ DeepTrust – Explainable AI Deepfake Detection

Overview

DeepTrust is an explainable AI system for detecting AI-generated images. It focuses on accuracy, transparency, and long-term learning using human feedback.

Why DeepTrust?

• Uses verified research-grade pretrained models
• No black-box APIs
• Visual explanations (heatmaps)
• Trust Score for user clarity
• Designed for long-term deployment

How It Works

1. Upload an image
2. Pretrained deepfake model analyzes artifacts
3. Trust Score (0–100) is generated
4. Heatmap shows suspicious regions
5. User feedback improves the model

Model Loading (Important)

DeepTrust uses Torch Hub to automatically load a verified pretrained deepfake detection model on first run.

• No manual model download required
• Model is cached locally after first use
• Fully legal and reproducible

Tech Stack

• Python
• Streamlit
• PyTorch
• EfficientNet
• SQLite

Future Scope

• Video deepfake detection
• Face-level analysis
• Browser extension

Disclaimer

DeepTrust is a decision-support system and should not be used as the sole authority for forensic conclusions.