Abstract: In the ever-evolving fields of computer vision and machine learning, real-time object detection presents a significant challenge. Our project dives into this domain, harnessing cutting-edge machine learning algorithms. By leveraging advanced methodologies like YOLO variants and efficient backbone architectures, our system aims to redefine real-time object detection. Traditional methods often compromise either accuracy or speed, limiting their effectiveness in dynamic environments. Our approach seeks the ideal balance, achieving both precise and swift object identification in complex scenarios. This integration of state-of-the-art techniques empowers our system to serve diverse sectors, from autonomous systems and security to interactive technologies. Through this endeavour, we envision a future where intelligent visual perception sets new standards for real-time object detection.

Keywords: Object detection, YOLOv7, Machine Learning, Faster R-CNN.

| DOI: 10.17148/IARJSET.2023.10847