Blind Vision Shirt

Project Title

Blind Vision Shirt

Project Area of Specialization Artificial IntelligenceProject Summary

Blind Vision Shirt is a hardware component that will make a new sense for blind people and to increase their “umwelt” that will wearable like normal shirt contain some vibratory motors and that should facilitate all blind people to see this whole universe through the vibratory pattern. This project aims to build a hardware application that will help blind people to analyze, study judge, remember visual information through the vibratory pattern. Those vibrations will help to understand and distinguish between different objects like numbers, characters (English), faces and such facets those will help to navigate blind people. Even blind will able to judge people and distinguish between them. The BVS will help blind people to read our normal life books, text messages, letters, currency, utility bills or even any kind of written patterns or set of characters of English Language. That device is an initial level to provide the cheapest device for blind people that can easily help them to AI object detection, face recognition, voice navigation, face detection and an OCR detection model in a single device.    

Problem Statement

There are several projects but no such project to help blind people to read books, newspapers, remember faces and all of the object information’s throughout their body by feeling vibratory patterns and the first time we will develop such hardware component to process that idea on blind people.

Project Objectives

This project aims to build a hardware application that will help blind people to analyze, study judge, remember visual information in a new manner. That will make a new sense for blind people and will able to prove that humans could judge this word with different manners. Those vibrations will help to understand and distinguish between different objects numbers, characters, faces or such facets those will help to navigate blind people. Even blind will able to judge people and distinguish between them.

Project Implementation Method

The BVS will be placed on the “Raspberry Pi Model 3 B”, a camera module and five coin vibratory motors embedded with Raspberry Pi. A blind person can command and operate hardware with voice commands so mic and sound module for Raspberry Pi is also embedded with that; for machine learning, we are going to use tensor flow. Tensor flow is Google's Machine learning framework that is opensource but for its training, we need a powerful machine containing a TPU or at least a GPU to develop and train models for our project. we required 2 laptops for data processing and object image labilizing data preprocessing and voice comment data development. Finally, a Powerfull GPU embedded computer with the Tensor flow framework   

Benefits of the Project

That project will help blind people to read our normal life books, text messages, letters, currency, utility bills or even any kind of written patterns or set of characters. Blind will able to remember faces with new manners. That is a new manner where we will provide an offline version of the trained model machine that will help blind people to help in analyzing.

Technical Details of Final Deliverable

The final deliverable product is a wearable shirt that contains sensors and a raspberry pi hardware component that is helpful for blind people and deaf people to judge this universe through vibrations as well as an android application that contains raspberry pi's trained models that will help blind people to judge objects using blind vision application. We are training our new models on Tensorflow Google's most powerful framework. Finally, we will develop and train our models and convert them into sensor flow lite and embed them in raspberry pi and android application. Our main task is to develop an offline version of the most efficient training AI robotic device that will help blind as well as deaf people to understand our world and change their lives. 

Final Deliverable of the Project HW/SW integrated systemType of Industry Education , IT , Medical , Others Technologies Artificial Intelligence(AI), Internet of Things (IoT), Robotics, Wearables and ImplantablesSustainable Development Goals Good Health and Well-Being for People, Life on LandRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	23970
Raspberry Pi 3 Model b+	Equipment	1	5520	5520
raspberry pi 3 camera	Equipment	1	800	800
GPU At least 2GB Nvidia or Gforce mini price	Equipment	1	8000	8000
transistors and wires	Equipment	1	150	150
Vibrating Mini Motor Disc with shild for Raspberry pi	Equipment	5	450	2250
16-channel 12-bit PWM / Servo-I2C Interface-PCA9685 for Arduino Raspbe	Equipment	1	250	250
Printed Tshirt with custom Junctions	Equipment	1	1000	1000
Other components and elements like OpenAI and online courses etc	Miscellaneous	1	6000	6000