Published Paper Details:

This project introduces a real-time object detection and localization system that leverages the YOLO (You Only Look Once) pre-trained model, integrated with an intuitive user interface developed using Python's Tkinter framework. The system aims to enhance object recognition by delivering immediate visual feedback along with audio-based alerts, announcing both the identity and relative position of detected objects within the camera frame. Users interact with the system via a graphical interface that features a "Start Detection" button, initiating a live video stream. The YOLO model processes this stream in real time, identifying objects and estimating their spatial locations--such as left, right, top, bottom, or center. To support accessibility, particularly for visually impaired users or in hands-free environments, the system incorporates text-to-speech synthesis to provide verbal descriptions of the detections. By combining efficient object detection, spatial localization, and audio output, this system demonstrates practical utility in areas such as surveillance, assistive technologies, and industrial automation. The integration of YOLO ensures high-speed, accurate detection, offering a seamless and responsive user experience.