Facial recognition at border control
Facial recognition technology is used at border control to verify a person’s identity.
This often involves people using an e‑gate, where the facial image obtained by the camera is compared to their scanned identity document.
The technology stores biometric information about individuals related to their face and facial expression. This information is used to make a facial map. The AI technology then verifies the person’s identity by comparing the data from the scan of the person’s face with the biometric data stored on the database and identity document.
What are the benefits of this technology?
Using facial recognition technology at border control is faster than traditional methods of identity verification, leading to shorter queue times. Research has found that some AI systems are more accurate than humans at confirming the identification of passengers, meaning that any security issues can be detected quickly.
What are the risks of this technology?
There are concerns that facial recognition technologies struggle to identify people with certain physical characteristics, which may lead to discriminatory practices at border control. For instance, research has found that facial recognition technologies often misidentify darker-skinned females, with error rates up to 34% higher when compared to lighter-skinned males, and do not work well for children, elderly people and non-binary people. The deployment of facial recognition software at border checkpoints also raises issues around consent for data use. Passengers are often given limited information about how to opt out of facial recognition in an airport or border setting.
What is machine learning?
Machine learning is a kind of AI. By learning from existing data (known as‘training data’), systems can identify patterns to make predictions or calculate probabilities. These systems are able to continuously adjust as they encounter more data. Based on what they learn from training data, they can perform a variety of functions, like playing chess, recognising faces or assessing welfare benefit applications.What is deep learning?
Deep learning is a subset of machine learning that uses neural networks with many layers – a method of processing data that has been modelled on the structure of the human brain, with many interconnected nodes organised in a layered structure. These neural networks are trained by automatically updating each of their layers as they are exposed to more data. Neural networks learn to make decisions and predictions through exposure to data, without being explicitly programmed to make specific decisions.What is biometric data?
In general, biometric data means data that relates to the physical characteristics of a person that can be measured, recorded and quantified, such as their face, fingerprints, palmprints, eye irises, etc. Biometric data can be used to identify a person and verify aspects about them, such as their age, gender or mood.What is facial recognition technology?
Facial recognition technology aims to identify or observe individuals by detecting the features associated with a human face. The technology analyses and measures distances between specific facial features and generates a unique representation (a ‘facial signature’) of each human face. This facial signature can then be compared against a database of stored images. Other types of facial recognition technology can assess the age of the individual whose face is being scanned or track their emotions and facial expressions. The technology can be used with the intent to uniquely identify individuals, but it does not always accurately identify individuals.