AI AND VIDEO ANALYTICS BLOG
Video Surveillance & Physical Security Industry Viewpoints
May 20th, 2020
Author: Stephanie Weagle

Efficient Contact Tracing with Video Content Analytics for Preventing the Spread of COVID-19

What is Contact Tracing

Public health experts are nearly unanimous in their opinion that contact tracing, alongside testing, is essential to quelling the COVID-19 crisis by breaking the chain of transmission. Testing indicates who is infected with the virus, while contact tracing enables others to learn if they have had contact with an infected person and take action accordingly.

WEBINAR: VIDEO CONTENT ANALYTICS FOR COMBATING THE PUBLIC HEALTH CRISIS

Naturally, people want and need to know if they have been exposed to a carrier; that way, they can self-quarantine to avoid potentially exposing their family, friends, and co-workers to the virus, and they can take action quickly if they have any symptoms. When someone is diagnosed as a carrier, it can be challenging for them and public health agencies to determine the people with whom the carrier had contact; one person – depending on his or her daily routine – may interact with dozens of people each day, and not be able to precisely recall his or her interactions.

Contact Tracing with Mobile & Geolocation Technologies

Technology can play an essential role in productively and rapidly determining which persons came into contact with an infected individual. For example, some companies such as Apple and Google are developing cell phone apps to trace those who are infected: When users voluntarily enroll in the app program and self-disclose when they have COVID-19, the cell phone can log the person’s movements and enable geolocation tracking, and can send a cell phone notification, or “ping,” when someone is or has been in the proximity of an infected individual. However, to date, these apps are limited in their effectiveness because they rely on mass adoption to drive significant results.

Facial Recognition & Appearance Similarity for Accelerating Contact Tracing

Given that many organizations already have CCTV infrastructure in place to enhance physical security, one technology that can drive efficient contact tracing is video content analytics. Video intelligence solutions are based on Deep Learning and Artificial Intelligence technology, and designed to detect all the objects in a video scene and then extract, identify, classify and index them for a variety of analytic applications – including object attribute and face recognition. Video content analytics is a reliable technology that can assist in contact tracing, while protecting the privacy of infected individuals. Whether in a bustling hospital or an office building, if an employee has contracted COVID-19 and discloses that information to his/her employer, the manager can use intelligent video surveillance to map out the individual’s movements and interactions, with a workspace. Using facial recognition or appearance attributes such as clothing color, the video surveillance operators can pinpoint the individual’s appearances across video feeds over the course of the incubation period. By filtering video, the surveillance operator limits the amount of video to review only to the instances where the affected induvial appeared, and thus saves hours of time in manual review. By reviewing the video segments where the affected individual appeared, the operator can understand which people that employee came into contact with and notify those individuals that they have been in contact with someone with Coronavirus and recommend self-isolating.

In short, when an organization has video intelligence software with “in the wild” facial recognition, it can take the following steps to use it for COVID-19 contact tracing:

  1. Employee self-discloses that he or she is infected with COVID-19
  2. Employer obtains or uses a digital image (still photo or video image) of that employee and uploads that image into its video content analytics system for conducting a filtered search across footage for face matches for that employee
  3. Appearance attributes can also be utilized, coupled with face recognition or as a standalone filter to identify the diagnosed individual
  4. When face matches are identified, the operator reviews video to identify persons that encountered the affected individual. The task of identifying affected co-workers is easier if the company also has a digital image database of all its employees
  5. Keeping the affected individual anonymous, the employer can notify at-risk employees or clients/customers with whom an infected person has interacted, and recommends that they self-quarantine for 14 days

It is crucial to maintain health and safety in workplaces such as retail stores, warehouses, and hospitals. In this COVID-19 era, contact tracing is an important element of workplace safety, and video content analytics can play an important role in contact tracing efforts to mitigate the spread of the virus.