Detection of COVID-19 in Intelligent Building Occupancy Management

智能建筑占用管理中的 COVID-19 检测

• 批准号: 555212-2020
• 负责人: Granger, Eric
• 金额: $ 3.64万
• 依托单位: École de technologie supérieure
• 依托单位国家: 加拿大
• 项目类别: Alliance Grants
• 财政年份: 2020
• 资助国家: 加拿大
• 起止时间: 2020-01-01 至 2021-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=705765
• 关键词: Detection; COVID; 19; Intelligent; Building

Measurement of building occupant information is important for energy efficiency, comfort, health, productivity, and security management. Given the current global pandemic, Distech Controls Inc. seeks to automatically detect building occupants with symptoms of COVID-19, or in close proximity to one another, and thereby limit propagation of the COVID-19 virus. The main objective of this project is to develop compact privacy-preserving deep learning (DL) models for building occupancy measurement systems that allow detecting people with high fevers, and in close proximity to others, and thereby limit the propagation of COVID-19. These models will rely on low-resolution multimodal (RGB and thermal IR) cameras that are co-located on a wall, portal, or ceiling of a building to estimate the physical distance, density, and temperature of people in a room. Expanding on its building occupancy measurement systems, Distech seeks to develop cost-effective DL models for cross-modal person detection, counting, and re-identification. This project involves a cross-disciplinary team from ETS and Distech, and will allow to intensify the exchange of ideas and resources, and establish long-term collaborative links. Focusing on the design of DL models for visual recognition applications from low, this project will focus on state-of-the-art research. By focusing on the development of cost-effective deep learning (DL) models for COVID detection through the multi-modal fusion of lower-resolution RGB-IR sensors, we anticipate that this project will lead to innovative AI technologies. Significant findings of this research project will be disseminated in high caliber scientific journals and conferences, and integrated into Distech building management solutions. This project also offers the opportunity for the training of highly qualified personnel to face current and future challenges in areas of strategic interest.

建筑物居住者信息的测量对于能源效率、舒适度、健康、生产力和安全管理是重要的。 鉴于目前的全球大流行，Distech Controls Inc.旨在自动检测具有COVID-19症状的建筑物居住者，或彼此非常接近，从而限制COVID-19病毒的传播。该项目的主要目标是开发紧凑的隐私保护深度学习（DL）模型，用于构建占用测量系统，允许检测高烧患者，并与其他人非常接近，从而限制COVID-19的传播。这些模型将依赖于低分辨率的多模式（RGB和热红外）摄像头，这些摄像头位于建筑物的墙壁，门户或天花板上，以估计房间中人的物理距离，密度和温度。在其建筑物占用测量系统的基础上，Distech寻求开发具有成本效益的DL模型，用于跨模态人员检测，计数和重新识别。该项目涉及来自ETS和Distech的跨学科团队，并将加强思想和资源的交流，并建立长期的合作联系。本计画将专注于视觉辨识应用之深度学习模型的设计，从低层开始，进行最新的研究。通过专注于通过低分辨率RGB-IR传感器的多模态融合开发用于COVID检测的具有成本效益的深度学习（DL）模型，我们预计该项目将带来创新的人工智能技术。该研究项目的重大发现将在高水平的科学期刊和会议上传播，并集成到Distech建筑管理解决方案中。该项目还提供了培训高素质人员的机会，以应对战略利益领域当前和未来的挑战。