A new machine learning and artificial intelligence technique developed by U.S. Army Research Laboratory researchers generates a visible face image from a thermal image of a person's face taken at night.
Automatic face recognition at night using thermal cameras can help soldiers recognize that an individual is on a watch list, for example.
Conventional cameras cannot capture facial imagery at night without active illumination, which would give away the camera's position. This is not the case with thermal cameras that capture the heat signature naturally emanating from living skin tissue.
The new approach uses advanced domain adaptation techniques based on deep neural networks.
The team found that merging global information, such as the features from the entire face, and local information, such as eyes, enhanced the visible face image.
The new approach achieved better verification performance than a generative adversarial network-based approach.
From U.S. Army Research Laboratory
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA
No entries found