Abstract
A person’s ability to recognize familiar faces across a wide range of viewing conditions is one of the most impressive facets of human cognition. As shown
in Figure 1, it is easy to conclude, for a known individual, that each image in the set shows the same person (British Prime Minister David Cameron), despite a wide range of variation in viewing angle, physical appearance, camera and lighting. In fact, familiar face recognition performance is often at or near ceiling level, even when the images are of poor quality [1] or artificially distorted. [2] At first glance, the aptitude for familiar face recognition may suggest a similar level of expertise for the recognition of unfamiliar faces, thus the reliance on face-to-photo ID for identity verification. [3] This is not the case, as recent research
shows people are surprisingly poor at recognizing new instances of an unfamiliar person. The poor recognition of unfamiliar faces is a concern for the United States. Many preliminary screenings involve facial recognition by security agents. In order for this method to be effective, more robust training for security
agents needs to be established. The Department of Defense utilizes facial and iris recognition technologies in order to eliminate human error in identifying persons of interest during surveillance operations. [4] DoD guidelines should be implemented by security agent guidance programs to ensure best practices in identification of potential threats.
in Figure 1, it is easy to conclude, for a known individual, that each image in the set shows the same person (British Prime Minister David Cameron), despite a wide range of variation in viewing angle, physical appearance, camera and lighting. In fact, familiar face recognition performance is often at or near ceiling level, even when the images are of poor quality [1] or artificially distorted. [2] At first glance, the aptitude for familiar face recognition may suggest a similar level of expertise for the recognition of unfamiliar faces, thus the reliance on face-to-photo ID for identity verification. [3] This is not the case, as recent research
shows people are surprisingly poor at recognizing new instances of an unfamiliar person. The poor recognition of unfamiliar faces is a concern for the United States. Many preliminary screenings involve facial recognition by security agents. In order for this method to be effective, more robust training for security
agents needs to be established. The Department of Defense utilizes facial and iris recognition technologies in order to eliminate human error in identifying persons of interest during surveillance operations. [4] DoD guidelines should be implemented by security agent guidance programs to ensure best practices in identification of potential threats.
| Original language | English |
|---|---|
| Pages (from-to) | 14-21 |
| Journal | The Journal of the Homeland Defense and Security Information Analysis Center |
| Volume | 3 |
| Issue number | 1 |
| Publication status | Published - 21 Mar 2016 |
Bibliographical note
© 2016, The publisher. Uploaded with permission of the publisher/copyright holder. Further copying may not be permitted; contact the publisher for detailsAbout this publication: The Journal of the Homeland Defense and Security Information Analysis Center (HDIAC) is published quarterly by the HDIAC staff. HDIAC is a DoD sponsored Information Analysis Center (IAC) with policy oversight provided by the Assistant Secretary of Defense for Research and Engineering (ASD (R&E)), and administratively managed by the Defense Technical Information Center (DTIC). HDIAC is operated by Information International Associates (IIa) in Oak Ridge, Tenn. Reference herein to any specific commercial products, processes or services by trade name, trademark, manufacturer or otherwise, does not necessarily constitute or imply its endorsement, recommendation or favoring by the United States government or HDIAC. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States government or HDIAC, and shall not be used for advertising or product endorsement purposes.
Copyright 2016 by IIa. This Journal was developed by IIa, under HDIAC contract FA8075-13-D-0001. The government has unlimited free use of and access to this publication and its contents, in both print and electronic versions. Subject to the rights of the government, this document (print and electronic versions) and the contents contained within it are protected by U.S. copyright law and may not be copies, automated, resold or redistributed to multiple users without the written permission of HDIAC. If automation of the technical content for other than personal use, or for multiple simultaneous user access to the journal, is desired, please contact HDIAC at 865.535.0088 for written approval.
Distribution Statement A: Approved for public release; distribution
is unlimited.
Keywords
- FACE RECOGNITION
- Face matching
- Biometrics