Abstract

According To The World Health Organization (Who), A Significant Number Of People Worldwide Experience Visual Impairments, Which Can Have Profound Personal Implications As A Vast Majority Of Our Perception, Learning, And Daily Activities Rely On Vision. Consequently, Individuals with Visual Disabilities Encounter Challenges In Education And Professional Settings Due To Inaccessible Technologies And Barriers To Basic Tasks In Daily Life. To Address This Issue, We Introduce A Speech-Based Assistive System That Enables Visually Impaired And Blind Individuals To Interact With Chatgpt Through Natural Conversation. This System Integrates Automatic Speech Recognition (Asr), Text-To-Speech (Tts), And Ai-Iot To Create A Seamless Interface With Chatgpt. Through Initial Hardware Designs, We Have Demonstrated The Potential Of This System As An Assistive Technology. The Project Utilizes A Custom-Designed Chatgpt Hardware System To Provide Assistance To Individuals With Disabilities. A Microphone Is Used To Capture The Voice Input, Which Is Then Sent To The Server. In Our Server, The Chatgpt Api Processes The Message, Searches For The Necessary Response, And Displays The Message. The Response Message Is Sent To A Voice Processor Module To Convert It Into Speech, Which Is Then Played Through A Speaker Module, Allowing The Individual To Hear The Result In Voice Form And Respond Accordingly. Additionally, A Slave Microcontroller Is Employed To Receive Command Details From The Chatgpt Processor And Control A Display, Indicating Course Changes And Displaying Start And Stop Statuses.