Unmask: Ensuring Media Authenticity in the Age of Deepfakes

Tagline: 🔍 Ensuring Media Authenticity in the Age of Deepfakes

Overview

Unmask is a cutting-edge service that combines advanced AI for deepfake detection with blockchain technology to provide tamper-proof certification of media authenticity. Users can upload media files (images, videos, audio), and the system will analyze them for signs of manipulation. If the media is verified as authentic, a digital certificate of authenticity is generated and stored on the blockchain. This certification can be accessed via an intuitive web application, and users can pay for the service using cryptocurrency.

Problem It Solves

In an era where synthetic media and deepfake technology are becoming increasingly prevalent, distinguishing between real and manipulated visual content is a critical challenge. Unmask addresses this issue by providing a reliable and accessible tool for detecting deepfakes and issuing verifiable certificates. This service empowers users to make informed decisions about the credibility of multimedia content, thereby enhancing trust in digital media.

Use Cases

1. Social Media Verification: Verify the authenticity of media by forwarding it to our WhatsApp bot or uploading it on our website.
2. Journalism and News: Journalists and news organizations can authenticate media before publishing to maintain credibility.
3. Legal Evidence: Legal professionals can verify digital evidence for admissibility in court.
4. Personal Use: Individuals can validate personal photos or videos used in online dating or professional profiles.

Key Features

• Deepfake Detection AI: Uses AI models (CNNs, RNNs, transformers) to detect deepfakes and provide a confidence score.
• Blockchain Certification: Uses Alephium for storing tamper-proof certificates.
• Cryptocurrency Payment System: Integrates Alphium for payment, enabling secure transactions.
• User Interface: Web-based application for easy media upload and certificate access.

Technologies Used

• Design: Figma
• Frontend: React.js (JavaScript)
• Backend: FastAPI (Python)
• Database: MongoDB
• Blockchain: Alephium
• AI Models: TensorFlow, PyTorch

Challenges We Encountered

During development, we encountered a version dependency error when installing requirements for the deep learning model. This issue was resolved by creating two separate requirements files to manage dependencies effectively.

---

Installation and Setup

To run Unmask, follow the instructions below.

Prerequisites

• Node.js
• Python 3.x
• MongoDB
• Alephium CLI
• Uvicorn

Clone the Repository

git clone https://github.com/iDhanush/AlephiumHackathon2024.git
cd unmask

Blockchain Node Server

Navigate to the alephium directory:

cd alephium

Install dependencies:

npm install

Compile the contract:

npx @alephium/cli@latest compile

Run the main server:

npx ts-node src/main.ts

Backend

Navigate to the backend directory:

cd backend33

Install Python dependencies:

pip install -r requirements.txt

Start the backend server:

uvicorn main:app --reload

Frontend

Navigate to the frontend directory:

cd frontend

Install frontend dependencies:

npm install

Run the frontend server:

npm run dev