Enhanced Spatial Awareness Enhanced Through Smart AI Glasses

The assistive technology, namely smart glasses, for visually impaired users. It effectively uses object detection and audio feedback in a novel manner. This thesis focuses on the enhancement of brand awareness through real-time audio cues concerning the surroundings for visually impaired users. There is a fairly straightforward design comprising of the Mounting of Pre-Trained YOLO11n for Object Detection and Camera on the Glasses. It is trained on the COCO dataset which contains various classes of objects making it suitable for real-world scenarios. Detected objects are further analyzed to develop audio feedback via speech synthesis. YOLO11n displayed effective performance in object detection and tagging in a controlled environment. The objects were detected and announced some tagged labels like "person," "chair," and "car." Confidence for objects like "bottle," "person," "cell phone," and "keyboard" was found to be 0.53, 0.66, 0.70, and 0.55, respectively. Preliminary tests have shown promising results in regard to the accuracy and speed. Sensor performance is hindered by ambient light changing and individuals intruding in the open in front. The project has brought forth the right kind of solution from the domain of assistive technologies for the blind. This project emphasizes the integration of computer vision and audio feedback that provides the disabled mobility and situational awareness. Future work shall include optimizing models for real-time practical applications, introduced diversity in recognized object types, and improvements to model’s stability in highly varying environmental conditions.