The Application and Ethics of Face Recognition. Q&A with Elke Oberg of Cognitec

Earlier this month, the face recognition company, Cognitec, revealed that it was working with RDW (the Netherlands Vehicle Authority), to provide a digital application platform for the renewal of drivers’ licenses. With some many stories about the use of face recognition making national headlines, the ID Bulletin spoke to Cognitec’s Marketing Manager, Elke Oberg, to cut through some of the hype and misconceptions, in order to understand the application of the technology and address some of the ethical issues.

What identification processes best suit face recognition?

The most reliable face recognition applications involve high-quality images taken in a controlled environment. Any photographs taken for ID documents, with their quality standards set by biometric image requirements, will yield high accuracy rates for face matching. Therefore, ID agencies are successfully using the technology for comparing a new ID photo to huge image databases to prevent fraud and data entry errors. Similarly, automated border control procedures involve ideal facial images, with the image on the passport chip being checked against live images from a video camera. Here, the person’s face can be perfectly positioned and lit, producing an optimal image for high matching rates.

When using face recognition for identification purposes, is it or can it be used as a single factor, or does it work best as part of a multi-factor authentication approach?

Most passports or ID cards only store one biometric, traditionally a facial image. Consequently, the database contains just a single biometric token for this person. In most cases, one modality will sufficiently serve the task at hand. But as we know, each biometric technology brings its own challenges to an identification process. Face recognition accuracy may decline as the person ages, wears too much makeup, thick glasses or headdresses. Therefore, combining two or more biometric modalities will certainly result in lower error rates and a more secure process.

There are a lot of organizations promoting face recognition solutions. What makes Cognitec stand out from the crowd?

Cognitec is the only company worldwide that has worked exclusively on face recognition technology since its inception in 2002. The company was never sold, acquired or merged, but has grown at a steady, responsible pace. As the industry landscape is changing quickly, the company is proud to maintain a stable, market-leading position within the industry, and to uphold the trust it has gained as the reliable, most experienced provider of face recognition technologies for enterprise and government customers worldwide.

There is a lot of debate surrounding the ethical use of face recognition in areas such as surveillance. Do you think these concerns are justified?

Being watched and tracked in a covert way goes against human rights in a free society. Unfortunately, each week the press reveals more instances of face recognition technologies misused, or used without transparency. In addition, governments and companies seem unable to protect large databases with personal data from hackers and cybercrime. Such questionable practices fuel the general fear of losing the right to privacy and other basic rights.

It is definitely necessary to observe the technical advances and deployments of face recognition technologies, and regulate their use in a balanced way to serve all members of society equally and ethically. All free countries should be involved in the privacy debate, hopefully to unravel the confusions around data consent and ownership. Finding the right use cases, where the consent is given in person and in a general sense, will ensure a healthy future of face recognition technologies in our society.

Do you think there is a lot of misunderstanding/misrepresentation of face recognition in media?

Unfortunately, many articles are using a rather sensational approach to point out the negative sides of using the technology. Statistics are often misinterpreted, or come from unscientific studies. The press focuses on the use of face recognition for mass surveillance, rather than exploring its outstanding capabilities to assist crime solving. Finding a person among millions of records is humanly impossible. An automated face recognition search yields results in mere seconds, and can move a life-saving investigation in the right direction.

For more information about face recognition and Cognitec visit: www.cognitec.com