Design and Development of an Integrated Internet of Audio and Video Sensors for COVID-19 Coughing and Sneezing Recognition
Objective
The objective of this research is to increase event recognition reliability using multi-audio and visual sensor integration by removing false positives.
Description
This research focuses on developing a scalable, real-time sensing solution to enhance situational awareness and early warning systems for the COVID-19 pandemic by recognizing risky behaviors, such as coughing and sneezing, associated with virus transmission. While traditional methods using audio-only or video-only sensors combined with Deep Learning (DL) algorithms have been employed for event recognition, these approaches often suffer from false detections and recognition failures. To address these challenges, sensor integration is proposed as a method to enhance precision and reliability in event recognition. Leveraging the widespread availability of low-cost audio and video sensors, this study introduces a real-time Internet of Things (IoT) architecture that integrates edge and cloud computing to improve the accuracy of coughing and sneezing recognition systems. This approach aims to deliver an effective, scalable solution for monitoring and mitigating COVID-19 spread.
Project Details
Collaborator(s): Sina Kiaei, Sepehr Honarparvar, Dr. Sara Saeedi, Dr. Steve Liang
Highlights:
Date: 2019 – 2021