Thermal face recognition over time

We present a comparative study of face recognition performance with visible and thermal infrared imagery, emphasizing the influence of time-lapse between enrollment and testing images. Most previous research in this area, with few exceptions, focused on results obtained when enrollment and testing images were acquired in the same session. We show that the performance difference between visible and thermal recognition in a time-lapse scenario is smaller than previously believed, and in fact is not statistically significant on existing data sets.