Army Trains AI to Identify Faces in the Dark

IEEE Spectrum  March 9, 2021
To develop a nighttime and low-light face recognition capability for the unconstrained or difficult lighting settings a team of researchers in the US (West Virginia University, Army Research Laboratory, Johns Hopkins University, University of Nebraska, industry) unveiled a dataset called Research Laboratory Visible-Thermal Face Dataset (ARLVTF) with over 500,000 images from 395 subjects. The data was captured using a LWIR camera mounted alongside a stereo setup of three visible spectrum cameras. Variability in expressions, pose, and eyewear were systematically recorded. The dataset has been curated with extensive annotations, metadata, and standardized protocols for evaluation. The researchers present extensive benchmark results and analysis on thermal face landmark detection and thermal-to-visible face verification by evaluating state-of-the-art models on the ARL-VTF dataset…read more. Open Access TECHNICAL ARTICLE

Posted in Uncategorized and tagged , , .

Leave a Reply