New computer vision and biometrics applications from AnyVision deployed in COVID-19 response effort
The comprehensive explorations of ways to combat the COVID-19 outbreak by Israel’s Defense Ministry has drawn in the computer vision capabilities of AnyVision, repurposing the deep-learning expertise the company used to develop its biometric facial recognition to detect COVID-19 cells, The New York Times writes.
In tests at Tel Aviv-Sourasky Medical Center, AnyVision is used on microscopic images to spot cells diverted from their usual function by the virus. Results can be delivered in a few minutes, and the false-positive rate may be five percent or less. A positive result would be followed up with another test, which takes longer, but the number of people taking the existing test could in theory by significantly reduced.
AnyVision technology is also used in hospitals, including the Sheba Medical Center, where it detects people entering a department without a mask from images supplied by 600 cameras in public areas, according to the report. Similar technology was also recently deployed to Tel Aviv-Souasky which also differentiates between different causes of high body temperature.
The system at Sheba Medical Center also provides information on anyone a hospital worker testing positive has had contact with based on facial recognition, as well as information about the distance and duration of their interaction.
A company blog post discusses the new nature and perception of safety in light of the pandemic.
Against the backdrop of changed social patterns, hesitancy to touch surfaces, and economic uncertainty, AnyVision asks how its technology can help reduce the spread of the deadly virus, how it can mitigate risk during economic stress, and how it can help people receive the services they need.
To those ends, the company plans to apply its technology to physical access control, detection of threats beyond human capabilities, and providing customer devices to businesses to support service delivery.
A whitepaper by AnyVision on “The Case for Scaling Computer Vision: Enhancing Safety in Uncertain Times” is offered free from its website, and the company held a webinar on a similar theme on April 30.
“Computer vision, the most powerful application of AI, has had a controversial past. As with every breakthrough technology, computer vision has been met with equal parts fear, curiosity and excitement,” the company writes in the post. “The ethics and value of every technology are judged by how it is applied, and the world has an opportunity – a mandate – to rapidly apply computer vision in ways that preserve privacy while making people safer. We see parallels in medicine: the world has unified to rapidly research, develop and deploy new medical technology in order to save lives. This sense of focus, urgency and collaboration has prevailed over historical bureaucracy, national borders and partisan lines.”