Skip to content

anshu2k24/VIPchakra

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

VIPchakra

VIPchakra is a solution that protects Very Important Persons (VIPs) from online threats and misinformation by using a full-stack application and machine learning.

Tech Stack

  1. Frontend: React, Tailwind CSS
  2. Backend: Node.js, Express (MERN stack)
  3. Machine Learning / AI:
    • Python
    • YOLO (object detection)
    • DeepFace (face recognition)
    • pHash (image hashing / similarity detection)
    • Sentence-BERT (semantic similarity / NLP embeddings)
    • Regex & Named Entity Recognition (NER) with spaCy

How to Run

Start the Frontend

cd frontend
npm run dev

Start the Backend

cd backend
node server.js

Start the ML Services

cd ml_services
uvicorn app:app --reload --port 8001

VIP Registration and Dashboard

  • VIP Registration: A VIP registers using an OTP-based authentication process.
  • Secure Details: The VIP provides sensitive details and 3–5 high-quality images. This information is stored securely to enable the system's monitoring capabilities.
  • Personalized Dashboard: Each registered VIP receives their own dashboard that provides a real-time feed of all flagged content and threats.

Core Features

  • Reused Content & Old News Detection: Identifies if old images or posts are being reposted by comparing content embeddings to prevent the spread of stale or misleading news.
  • VIP Recognition: Detects VIPs in images using face recognition. The system can also flag multiple VIPs if found in the same content or image.
  • Sensitive Information Monitoring: Flags content containing VIP-sensitive information leaks, wrong news, or direct threats.
  • Real-Time Monitoring: The system continuously monitors posts on platforms like x.news. Any flagged content is immediately reflected on the VIP's dashboard.
  • Post Flagging: Automatically flags suspicious posts and provides a clear reason for the flagging.
  • User Flagging: Flags user accounts that consistently share problematic content.
  • Manual Report Button: Allows users to report posts flagged incorrectly by the model.
  • Email Notifications: The system sends an immediate email notification whenever flagged content or images related to a registered VIP are detected.

Team

  • Anshu P: Backend, ML services, threat detection, data analysis, overall architecture
  • Dhruva Kr: Frontend, UI/UX, dashboard, visualization

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors