FNA.ai (Fake News Analyzer) is an innovative platform designed to combat misinformation. By harnessing cutting-edge technology, we ensure the authenticity of news content across images, videos, and textual formats. Our solution integrates AI and blockchain to provide scalable, secure, and immutable verification processes.
Our platform employs advanced AI models to analyze and flag tampered or manipulated media, safeguarding the authenticity of uploaded content.
Dataset Used: The model is trained on the Fake AVCeleb dataset, ensuring high accuracy in identifying deepfake content.
- Input: Upload news content in image or video format.
- Processing: The platform generates concise summaries using FastAPI.
- Output: A clear and digestible summary is provided for quick understanding.
π FastAPI Summarization Code: GitHub Repository
- Summarized content is compared with trusted news websites.
- If the description matches verified sources, the news is marked real; otherwise, it is flagged as fake.
- Base64 Encoding: Verified news is converted into Base64 format.
- IPFS Storage: The encoded data is stored on IPFS for decentralized, scalable storage.
- Pinata Integration: Ensures persistent storage of IPFS content.
- NFT Creation: Converts verified content into NFTs (Non-Fungible Tokens).
- Polygon Blockchain: Stores NFTs securely, ensuring immutability and scalability.
- Smart Contracts: Automate the verification and validation process via Polygonβs PoS mechanism.
π Node.js Blockchain Code: GitHub Repository
- AI-Powered Deepfake Detection: Detects and flags tampered media.
- Efficient Summarization: Generates summaries for better readability.
- Trusted Verification: Matches news content with verified sources.
- Secure Decentralized Storage: IPFS & Pinata ensure data integrity.
- Immutable Proof: Polygon Blockchain ensures transparency & tamper-proof records.
- Web3-Enabled Trust Badge: Verified news receives a blockchain-backed credibility badge.
| Category | Technologies Used |
|---|---|
| Deepfake Detection | TensorFlow, OpenCV |
| News Summarization | FastAPI |
| Blockchain | Node.js, Polygon, Solidity, Smart Contracts |
| Decentralized Storage | IPFS, Pinata |
| Front-End Framework | React |
| API Integrations | NewsAPI, YouTube API |
- Clone the Repository:
git clone https://github.com/Priyank911/FNA.ai.git
- Set Up the Environment:
- Install dependencies for each module as per their documentation.
- Obtain API keys for NewsAPI, YouTube API, and Pinata.
- Run the Services:
- Summarization: Navigate to the FastAPI folder and start the service.
- Blockchain Integration: Navigate to the Node.js folder and start the backend.
- Real-time AI analysis with enhanced deepfake detection models.
- Decentralized Fact-Checking Community leveraging DAOs on Polygon.
- Multi-language Support for diverse misinformation detection.
- Enhanced UI/UX for a seamless, interactive experience.
We encourage contributions to improve FNA.ai! Follow the standard GitHub Flow for submitting issues and pull requests.
This project is licensed under the MIT License. See the LICENSE file for details.
- Fake AVCeleb Dataset: Training the deepfake detection model.
- IPFS & Pinata: Providing decentralized storage solutions.
- Polygon Blockchain: Ensuring scalable, trustless verification.
- FastAPI: Enabling efficient backend processing.
- TensorFlow: Powering AI-driven analysis.
