Design of Hardware Interfacing with ChatGPT as an Inference Engine for Visually Impaired
S. P. Vimal, V. Kiruthika, M Hariharan, S. Midhilesh, G. Kabilan · 2024
This study proposes the design and implementation of a hardware interface system that leverages ChatGPT, a powerful natural language processing model, as an inference engine to assist visually impaired individuals in accessing information and performing everyday tasks. The proposed system aims to provide an intuitive and user-friendly interface that enables visually impaired users to interact with ChatGPT using voice commands and receive spoken responses, thereby enhancing their accessibility to digital content and services. The hardware interface consists of a portable device equipped with a microphone for input and a speaker for output, along with additional tactile and auditory feedback mechanisms to facilitate user interaction. The proposed system architecture integrates ChatGPT’s natural language understanding capabilities with the hardware interface, enabling seamless communication between the user and the AI model.