YoLoV8 Meets the Cloud: A High Performance Real-Time Detection Framework

Dr. J. Jeyshri

Abstract: This paper presents the development and implementation of a real - time object detection system using the YOLOv8 deep learning model integrated with a cloud - based environment, specifically Google Colab. The system is designed to process both images and video streams, offering accurate detection capabilities with minimal computational overhead. By leveraging cloud infrastructure, the model bypasses local hardware limitations, making high - speed and high - accuracy object detection accessible and scalable. The proposed architecture incorporates OpenCV for preprocessing, YOLOv8 for inference, Supervision for result visualization, and Roboflow for efficient model management. This modular, cloud - centric design demonstrates practicality for research, education, and prototype development.

Keywords: YOLOv8, Real - Time Object Detection, Google Colab, Deep Learning, Cloud Computing, Roboflow, OpenCV, Supervision, Computer Vision