- ryanjenkins9
- Sep 25, 2024
- 7 min read
Posted: 31 October 2022
Updated: 25 September 2024
School of Human Sciences
Institute of Lifecourse Development
University of Greenwich
London SE10 9LS
www.superrecognisers.com
super-recognisers@greenwich.ac.uk
Twitter: @GRecognisers
Brand new project: Identification accuracy when faces have aged
Professor Josh P Davis (University of Greenwich)
Please donate 8-12 photos of you, taken in different years, using an updated easy, quick, photo upload system. In return, we will send a £5 GBP Amazon voucher.
Once super-recognisers have seen a face or faces in a test, they will probably find second attempts on that test to be too easy. Therefore, we need a flow of new facial images to include in updated versions of our current tests, and at the moment we are attempting to create new Age progressed Tests, designed to test the limits of human capabilities when images of the same person or different people vary in age.
For your information, every week, a team of volunteers search through our facial images, to find those that might be mistaken for one another. These "different face" image pairs, together with some "same face" image pairs are pilot tested, often more than once, and some are then included in our new or updated tests. This process takes at least a year.
To follow best practice ethical procedures almost all images have been donated by Greenwich Research Volunteer Pool members and visitors to our website. We believe that volunteers and visitors are the most informed about what we do. When asked to provide informed consent for the use of those images in different research, these donors will understand exactly what their images will be used for. This is unlike some face recognition technology creators who without asking for anyone' permission, illegally trawled the internet (i.e. Facebook, Instagram) to access billions of facial images.
Summary: Passport officers, law enforcement, genealogists, and many other professions are often tasked with deciding whether 2 or more photos depict the same person or not. Photos may show people taken at different times in their life (e.g., passports - up to 10-years; genealogists - sometimes longer). Sometimes, those depicted are no longer alive (e.g., genealogists, police), and sometimes being certain that two images do, or do not depict the same person, may help in providing more evidence, or ruling out one of those depicted as being involved in a crime (e.g., police).
Forensic Facial Comparison Experts may be asked to apply their techniques to such photos, and if DNA or fingerprints are available, then accuracy of identification will of course be high. However, resources may be limited, and such methods are costly and take time. Rapid decisions may sometimes be critical and those making such decisions may also be the gatekeepers to decisions about resource allocation.
In a series of ongoing studies, this project will aim to evaluate human accuracy at this task and to measure the impact on accuracy, speed, and confidence in decisions from: -
Variations in the ages of the person/people depicted in the two photos.
Variations in the perceived facial distinctiveness/typicalness of those depicted in the two photos.
Variations in the age of the photo of the younger person, when the age interval between the two photos remains constant.
Variations from non-aging related changes impacting the appearance of the older person in the two photos.
Variations in the age(s) of research participants and the age(s) of those depicted in the two photos.
Variations in the face recognition ability of participants contributing to the research.
Variations in image quality impacting appearance of the person/people depicted in the two photos.
Variations in susceptibility to potentially biasing information on such judgements. For instance:
A witness, who turns out later to have played a part in the crime, states with certainty that one of the slightly blurred photos depicts a person who is known to be in the second photo, in order to draw suspicion away from them.
An honest, but mistaken witness views the most recent photo and states incorrectly that it is not the person known to be in the older photo.
Face Recognition Technology states probability of a match = 90% or 50% or 10% to the same two photos.
Photos may also be retained in new tests of these abilities long after the basic research is finished.
However, to achieve our goals we need more photos please, and to avoid any potential ethical criticisms, only adults of 18-years and above, who understand to what purpose their photos may be used for, can properly provide informed consent (see below). This is one of the reasons we try to avoid using databases of photos assembled by others.
We find that people from a UK or European heritage are most likely to contribute (probably because about 80% of the UK population had grandparents born in the UK), but we still need plenty more of these photos please.
However, we need far more photos depicting people from all cultural backgrounds please. We want to do our best to avoid potential criticisms of any real or perceived bias, and we can only achieve this with the help of people reading this notice.
Facial ID checks will be with us for years to come, and your contributions will hopefully help us achieve our long-term aims of assisting in reducing the risk of entirely innocent errors.
If you have provided photos to us in the past, please do not do so again. You will not be eligible for any more Amazon vouchers.
Please click below to upload your photos – you will find more ethics information.
More information below
We are asking volunteers to provide facial images of themselves for use in these tests. UK passports can be 10 years old, so this has practical applications in the real world.
We please need at least 8 images from each volunteer (maximum = 12), each taken in a different year in your life.
High-quality selfies or posed social media profile photos in which you are looking towards the camera are likely to be perfect.
If uploaded to social media, they are also probably date-stamped, and you will be able to work out your age at the time.
For example, if you are 40 today, photos showing you at the following ages would be perfect.
40 38 36 30 26 23 19 16
Or
40 39 38 37 36 35 34 33
As long as they are from different years, you will receive a £5 Amazon voucher. We are unable to provide vouchers in other currencies – sorry.
We often get asked why we do not use photos from free-to-use facial databases. One of the most important reasons is that anyone who takes our tests can be assured that the people depicted provided full informed consent. There has been a lot of publicity around some computerised face recognition research being conducted without this guarantee. We want to make sure we are following best research and ethical practice.
In addition, most people who provide images are volunteers who also take part in our research. As such, they will be fully informed about how their images will be used when providing consent. It is not possible to make this guarantee about any other photos we may wish to use.
At the bottom of this blog is information about ethics, GDPR, and privacy, and anonymity and how we protect participants’ rights if they supply such stimuli.
Thank you for your support. Josh Davis
Any questions please e-mail: super-recognisers@greenwich.ac.uk
Ethics, data protection and privacy
Professor Josh P Davis is a Chartered Psychologist and an Associate Fellow of the British Psychological Society (BPS) and is therefore bound to follow BPS ethics policy (https://www.bps.org.uk/news-and-policy/bps-code-ethics-and-conduct). He is also a member of staff of the University of Greenwich. All projects are approved by the University of Greenwich Research Ethics Committee (UREC) (https://www.gre.ac.uk/research/governance-and-awards/research-ethics-committee), although collaborations with other universities may first be approved by their ethics board and noted by UREC.
The photo upload procedure was first approved by the University of Greenwich Research Ethics Committee in 2015 (14.5.5.2).An amendment to procedures was approved by the same committee in June 2021.
All research data and images stored by the University of Greenwich are retained on a password protected database, and procedures are compliant with General Data Protection Regulation (GDPR) requirements, tailored by the Data Protection Act 2018 (https://ico.org.uk/for-organisations/guide-to-data-protection/guide-to-the-general-data-protection-regulation-gdpr/) and the University of Greenwich data protection policy (https://docs.gre.ac.uk/rep/vco/data-protection-policy).
All volunteers on the Face and Voice Recognition lab database have been invited to provide images (many as part of an invite to contribute to other research). We also have an invite on the website. Because there are so many people on the database there is virtually zero possibility of us identifying anyone who provides images.
Participants make up a personal code when uploading photos – we do not use their normal anonymous codes issued for our research.
If someone wants to receive the £5 voucher, they can enter their e-mail address at the end. However, this is in an entirely separate URL link that makes it impossible for us to match the e-mail address back to their photos. It is impossible to get to this step unless at least 8 images have been uploaded.
All images are anonymously coded on the database, linked only by a common code used to match the images supplied by the individual (e.g., someone’s 8 images could be stored as AA001_1, AA001_2, AA001_3 etc etc.).
We will retain age, gender, and ethnicity information against the image files for quick searching. For instance, we often want to find photos of two people from the same demographic background who might be mistaken for one another.
Images may be used in various online tests in future – there are different levels of consent (e.g., we sometimes show images in the media for educational purposes, and we ask for extra consent for this in advance).
If you spot your photo being used in one of our tests and you wish us to delete this, then please make a note of the test name and ideally trial number in that test and please contact super-recognisers@greenwich.ac.uk. If you remember the personal code you gave us with the photo it might help us identify it too.
Most of the other face recognition tests developed around the world display students. There are probably tens of thousands of research projects every year conducted by psychology students using such stimuli, let alone the hundreds, if not more, that are published. In other words, those who created them, know exactly who contributed. We believe that our processes provide far greater anonymity and privacy guarantees. As is noted in the instructions, “someone who knows you might recognise you. However, we are only interested in unfamiliar face recognition”.
We do not want to know the real identity of anyone who supplies us images – and if ever images get viewed by, for instance, police officers taking part in our research, they are fully aware that volunteers supplied the images – none are criminals.
Please ask questions at super-recognisers@greenwich.ac.uk if any of the above is unclear.
Comments