Remember meForgot password?
    Log in with Twitter

article imageFace recognition technology works at night

By Tim Sandle     Apr 28, 2018 in Technology
A new type of facial recognition technology has been developed to works in the dark, allowing night-time recognition and identification of a people.
The research comes from the . Here technologists have developed an artificial intelligence and machine learning platform that produces a visible face image based on a thermal image of a person's face. The image is captured in low-light or nighttime conditions. Given the success rate, the technology could lead to enhanced real-time biometrics. In terms of the intended military use, the technology could assist with post-mission forensic analysis stemming from covert nighttime operations.
There are potential civilian applications to. Face recognition is increasingly being used in crowded spaces, such as to replace a boarding pass and shorten time at customs or to identify terror suspects to prevent public threats. Facial recognition also has other applications, like improving social networks and the curating of photographs for news media.
READ MORE: 'Face recognition' for galaxies
To develop the new platform, special thermal cameras called Forward Looking Infrared sensors, were deployed on aerial and ground vehicles. These enable devices to digitally capture facial imagery for recognition without active illumination. Deep neural networks were then 'trained' to spot faces of subjects and to determine their identify. This worked by using a non-linear regression model to map a given thermal image into a corresponding visible latent representation, followed by an optimization problem that projects the latent projection back into the image space.
A conceptual illustration for thermal-to-visible synthesis for interoperability with existing visibl...
A conceptual illustration for thermal-to-visible synthesis for interoperability with existing visible-based facial recognition systems.
Eric Proctor, William Parks and Benjamin S. Riggan
According to one of the developers, Dr. Benjamin S. Riggan: "This technology enables matching between thermal face images and existing biometric face databases/watch lists that only contain visible face imagery. The technology provides a way for humans to visually compare visible and thermal facial imagery through thermal-to-visible face synthesis."
In further facial recognition news, Facebook has begun the process of asking its European and Canadian users to allow the social media giant to use facial recognition technology. This is for the purpose of identifying users in photos and videos. See the Digital Journal article "Facebook aims for facial recognition consent in EU and Canada."
More about Face recognition technology, Facial recognition, idenitity
Latest News
Top News