The architectural design of smart blind assistant using IoT with deep learning paradigm

Abstract

Abstract Machine learning and the Internet of things (IoT) play a significant role in digitizing the modern world. Deep learning in object detection leads to a sophisticated solution, and Virtual assistant can be favorable for visual impairment. This paper reflects an architectural design of smart blind assistant using the mechanism of deep learning embedded with IoT. The proposed model introduces an intelligent cap using the raspberry Pi and camera module, along with a deep learning paradigm. The proposed model presents a smart blind stick s structural design that utilizes a microcontroller with multiple sensors. The manuscript also provides a development process of virtual assistant that acts as a manager of complete integration. The model employs IoT and Bluetooth connectivity for instant data monitoring. The authorized person keeps watching on visual impairment using the IoT cloud server. To examine the proficiency of this anticipated model, object detection using deep learning, sensor data calculation, and system usability (SUS) are enumerated and interpreted. The SUS score of this work is 86%. However, the proposed system will be workable and handy in the daily activities of a blind person.