Home - PatilAntariksh/Mind-For-The-Blind GitHub Wiki

Project Overview

"Mind-For-The-Blind" is a help assist technology application made for the visually impaired individuals by providing help in the overall navigation , currency detection, and AI-powered conversational support. The mobile application integrates machine learning, natural language processing, and secure video communication to enhance the daily lives of blind users. The system consists of a mobile application, backend API, AI assistant, currency detection module, video call functionality, and cloud database authentication. Our major goal is to to empower visually impaired individuals to navigate their daily lives independently by providing an accessible, intuitive, and secure mobile application that assists with currency identification, Real time safe navigation, and voice-based guidance.

Project vision

To empower visually impaired individuals to navigate their daily lives independently by providing an accessible, intuitive, and secure mobile application that assists with currency identification, Real time safe navigation, and voice-based guidance. The vision of "Mind-For-The-Blind" is to create an accessible solution for the visually impaired that significantly lowers their dependency on their helpers and quality of life is improved.

We want to develop a wearable headset equipped with a high-definition camera that seamlessly connects with the mobile application. This headset will allow visually impaired users to experience real-time object detection, navigation support, and AI-powered assistance hands-free. The camera will continuously analyze surroundings, recognize objects, read text, and provide real-time audio feedback directly to the user. Additionally, the system will integrate with cloud-based AI models to offer instantaneous scene interpretation, improving mobility and awareness in various environments. This ambitious feature aims to create a fully immersive, real-time guidance system, setting a new benchmark for assistive technology. We also plant to add a signal detection model to help the visually impaired person cross the road.

Project Scope

Scope of the Project:

  • Currency Detection: Implement ML models to detect and identify currency denominations from user-scanned images.

  • Video-Call-Based Navigation: Develop a secure video-call feature to connect users with volunteers/navigators for real-time assistance.

  • Voice-Based UI Guidance: Integrate voice-based navigation and feedback to help users interact with the app.

  • AI-Powered Personal Assistant: Provide an intelligent voice assistant to answer user queries and guide them through the app.

  • Security Features: Ensure secure access and encryption for video calls and user data.

  • Cross-Platform Development: Use the Flutter framework to develop the app for both Android and iOS platforms.

Roles

  • Project Lead: Antariksh Sanjay Patil
  • Project Manager: Harsh Singh
  • Architect / Tech lead: Krinal Akbari, Antariksh Sanjay Patil
  • Developers: Antariksh Sanjay Patil, Krinal Akbari, Swaraj Bhalerao, Khushi Suman
  • DevOps / Automation: Khushi Suman, Krinal Akbari, Akhil G
  • Test Engineer: Swaraj Bhalerao
  • UX Designer: Khushi Suman, Akhil G
  • Requirements Engineer: Harsh Singh, Swaraj Bhalerao

Project Risks

  1. Technical Risks Machine Learning Model Accuracy Risk: The currency detection or AI assistant may predict wrong results. Voice Recognition Errors Risk: The AI assistant may struggle to correctly interpret verbal inputs such as in a disturbing environment.
  2. User Experience Risks Internet & Connectivity Issues Risk: The app relies on stable internet connections for video calls and AI assistance, which can be an issue. Also, smallest jitter in the video chat room may lead to life threating situation of the user in an event of road crossing.
  3. Operational & Deployment Risks High Computational Requirements Risk: Running AI models and video streaming may require high processing power, which may lead to low performance on devices with weak processor.

Assumptions

  • Smartphones with biometric authentication capabilities
  • Blind users will have some assistance setting up the app for the first time.
  • Users are willing to trust AI-based assistants