|
Direct Communications Unit
bcsl@blueyonder.co.uk Reference: T21200 8 14 October 2008
Dear David,
On a point of clarification, your email of the 21 August 2008 was deemed to be a request for information under the Freedom of Information Act and hence a formal reply to you was made through the Deputy Director of HOSDB. You make a number of points in your email. I understand that you have been separately in dialogue with the lead author of the feasibility study, Tony Mansfield, and I am sure that you will not want me to go over the material which he has discussed with you. Regarding one of the principal comments you make - that there is no reason to believe that face recognition 'technology now works better than it did five years ago' - I believe you are aware of the conclusions of the Face Recognition Vendor Test 2006 [1], which demonstrated that there has been considerable improvement. These results confirmed the feasibility study's observation that this technology could be applied successfully in a one-to-one (verification) mode. These tests were conducted by a consortium led by the US National Institute of Standards and Technology. Of course, it is only once this technology is tested in a specific context, with optimised algorithms and reference images, that we can be sure that such improvements translate into capabilities which can be deployed in a working system. Operational testing, e.g. in Australia and in Portugal, has confirmed the improvements which the NIST technology tests have identified. These tests took advantage of the higher quality reference images which are now available in electronic passports issued in the European Union and elsewhere, and whose specification has been set by an internationally developed standard, as we mentioned in our earlier reply. You quote from section 3.4 of the National Audit Office report on the introduction of ePassports which I see also notes that the Identity and Passport Service 'believes there is good potential in the future for one-to-one comparison of the image held on the passport chip with the passport holder standing at border control, which could eventually enable automated border control of the sort currently being trialled in Australia'. It is this potential for verifying one-to-one identity which the UK Border Agency (UK BA) is now evaluating at Manchester Airport. In the trials, we invite volunteer adult passengers who hold e-Passports from EEA countries to pass through the gates equipped with the face recognition system. In addition, a UK Border Force Officer will view the whole process on a monitoring station and will intervene if they have any suspicions. The supervising Border Officer will also stop a random proportion of passengers for an additional manual check. The performance data from the gates is being recorded (number of images taken, the match score, time to process each passenger, etc) and the anonymised images will be retained for analysis by independent experts so as to ascertain accuracy of the facial matching. We will use both these images and published academic sets to assess the accuracy of the technology used. There will also be a number of tests with human observers to see how their matching capabilities compare to those of the automated system. The trial at Manchester is scheduled to last six months and UK BA expects that the Automated Gates will be used by a significant number of eligible passengers. The resulting evaluation will be of the system as a whole, and not just a measure of the performance of the facial matching component. I would also like to take the opportunity to clear up one area of confusion regarding verification and identification accuracies. As you have observed, para 58 in our report does indeed state that 'For one-to-one identity verification, it is only necessary to use a single finger, a single iris or face recognition'. This is confirmed by the matching error rates as noted earlier in para 52, in the section on identification accuracy, and where we comment that such figures also show that face recognition on its own would indeed not be appropriate for 'one-to-many' matching in very large scale systems. You have also referred to the results of the Biometrics Enrolment Trial. Its objective, as set out in section 1.1.2, was 'to test the processes and record customer experience and attitude during the recording and verification of facial, iris and fingerprint biometrics, rather than test or develop the biometric technology itself. It was not a technology trial.' Performance figures have been published as part of the documentation, but it is misleading to use this data from an enrolment trial, especially in discussions on the verification performance to be expected from fully engineered modern systems. Finally, you refer to reservations regarding the maturity of biometric technologies which the Science and Technology Select Committee reported hearing from policy officials working for the US Department for Homeland Security. IPS had in-depth meetings in 2006 with several very senior officials responsible for the operation, development and management of US-VISIT who strongly rebutted the assertions made by the committee and pointed to the marked success of the technology employed in US-VISIT which allows rapid one-to-many matching of fingerprints on a database of 40-50 million individuals. I trust that these comments resolve your concerns. I am encouraged by your interest in the potential of biometric technologies in the safeguarding of people's identity and the privilege of UK citizenship, as well as in securing our borders and controlling migration for the benefit of our country. Regards
Marek Rejman-Greene
1. P. Jonathon Phillips et al, FRVT 2006 and ICE 2006 Large Scale Results, http://www.frvt.org/FRVT2006/docs/FRVT2006andICE2006LargeScaleReport.pdf (March 2007, accessed 2 October 2008)
|
|