Voice-activated Chatbot for Visually Impaired Users to Detect Objects and Navigate in Real-Time

This research aims to develop a speech-activated chatbot that enhances voice-driven navigation for visually impaired users by employing real-time recognition algorithms featuring a high operational accuracy of 92 percent. Two different approaches were used in the research. One strategy acknowledged the limitations of manual navigation and proposed a fix emphasizing voice-enabled navigation using Dialogflow making it effortless to use and open for people with impaired vision. The other strategy placed a high emphasis on accuracy, giving consideration to precise real-time object analysis and recognition. This study shows how to effectively integrate technical competence with user-centered design to overcome manual navigation and interaction. The existing concept with an ANN of 89 percentage was replaced with the proposed concept of 92 percentage for enhanced real-time object detection and shows a significance of about 0.021. In this work, it is observed that using YOLOv8 for real-time object detection has significantly better gain than using ANN.