Is video redaction software the answer to facial recognition concerns?
Facial recognition is trending. A biometric tracking technology that identifies and authenticates individuals via biometric data, it is already being used in general surveillance, fraud prevention systems and even check-in systems at airports. But companies, governments and regulators are becoming increasingly concerned with how to balance the convenience of facial recognition against inevitable privacy concerns.
How does facial recognition work?
A computer-based system, facial recognition technology may recognise or confirm a person from a digital image or video frame. It finds a match by taking a picture of a person's face and then using algorithms to compare it to a database of previously taken pictures.
Finding a face in a frame of a picture or video is the first step. The system locates and isolates the face region in the image, then extracts distinctive facial characteristics, including the distance between the eyes, the shape of the nose, the contours of the cheeks, and the jawline. As a result, "faceprints" – mathematical representations of these features – are created. To identify a match, the faceprints are next checked against a database of previously stored faceprints. The faceprints of each face in the database are compared to the similarity of each faceprint of the target face. The technology then recognises the individual in the picture or confirms that they’re not in the database - depending on if they are a match.
Why is live facial recognition problematic?
Live facial recognition (LFR) is when facial recognition technology is used in real-time, assessing a live camera feed of faces to potentially match faces to those of persons of interest held on file.
Even though some may argue this is useful in identifying potentially dangerous individuals moving across crowded spaces, it is often seen as a huge invasion of privacy.
The technology has been known to misidentify suspects, leading to wrongful arrests, and there are growing concerns that have led to calls to ban the technology entirely in the UK, the EU and beyond.
Balancing facial recognition technology and privacy
There are pressing questions about how individual privacy is prioritised when it comes to the deployment, use and management of (live) facial recognition. If data laws are not yet able to accommodate these advancing biometric technologies, regulators need to specify how data in these circumstances are protected. In the meantime, organisations should prioritise privacy and go above and beyond when using these - as some would say - invasive technologies. This should be a way to garner public trust and protect themselves against data breaches and possible hacks.
In order to comply with privacy laws and maintain the trust of customers and stakeholders, firms should use video redaction software, such as Secure Redact, to anonymise video footage appropriately before it is stored or shared.