A smart and scalable system is required to schedule various machine learning applications to control pandemics like COVID-19 using computing infrastructure provided by cloud and fog computing. This paper proposes a framework that considers the use case of smart office surveillance to monitor workplaces for detecting possible violations of COVID effectively. The proposed framework uses deep neural networks, fog computing and cloud computing to develop a scalable and time-sensitive infrastructure that can detect two major violations: wearing a mask and maintaining a minimum distance of 6 feet between employees in the office environment. The proposed framework is developed with the vision to integrate multiple machine learning applications and handle the computing infrastructures for pandemic applications. The proposed framework can be used by application developers for the rapid development of new applications based on the requirements and do not worry about scheduling. The proposed framework is tested for two independent applications and performed better than the traditional cloud environment in terms of latency and response time. The work done in this paper tries to bridge the gap between machine learning applications and their computing infrastructure for COVID-19.

中文翻译：

---

使用深度神经网络和云计算在工作场所进行智能 COVID 监控的可扩展框架

需要一个智能且可扩展的系统来安排各种机器学习应用程序，以使用云计算和雾计算提供的计算基础设施来控制像 COVID-19 这样的流行病。本文提出了一个框架，该框架考虑了智能办公室监控的用例，以监控工作场所以有效检测可能的 COVID 违规行为。拟议的框架使用深度神经网络、雾计算和云计算来开发可扩展且对时间敏感的基础设施，该基础设施可以检测两种主要违规行为：戴口罩和在办公环境中员工之间保持至少 6 英尺的距离。所提出的框架的开发愿景是集成多个机器学习应用程序并处理流行病应用程序的计算基础设施。所提出的框架可供应用程序开发人员根据需求快速开发新应用程序，而不必担心调度问题。所提出的框架针对两个独立的应用程序进行了测试，并且在延迟和响应时间方面表现优于传统的云环境。本文所做的工作试图弥合机器学习应用程序与其 COVID-19 计算基础设施之间的差距。