Craig Guthrie, editor
As one of the more emotive modalities in biometrics, the potential merits and downsides of facial recognition technology have long been a source of heated debate from outside and even within the biometrics industry.
For industry insiders, it seems that some journalists are very eager to jump on scare stories and cite potentially negative aspects, while ignoring the security benefits that help deter and capture terrorists, murderers and fraudsters.
Meanwhile, privacy advocates say that left unchecked, any invasive spread of face-based technology will lead to a dystopian future of oppressive surveillance. Who is the public to believe?
One constant amid the often tense debate is face recognition’s staying power: Given the technology’s versatility, accuracy and wide range of applications, there seems little doubt that we will also be seeing a lot more of it in the near future.
To delve deeper into the subject matter, Planet Biometrics has quizzed two experts from one of the leaders in face recognition tech, MorphoTrak.
In a free-ranging discussion, we sought the views of experienced biometrics industry professional and Senior Director of Strategic Marketing and Government Relations, Teresa Wu, and Rob Sprecher, MorphoTrak Program Director and lead on Law Enforcement Technology and Biometrics.
Craig Guthrie: In terms of facial recognition, what private-public collaboration models need to be considered for establishing best practices and guidance, while considering vital issues such as privacy?
Teresa Wu: I believe there is a lack of dialogue or collaboration right now. There are shared concerns, but there are very few collaborative avenues to address those concerns. I do observe that a lot of conferences now have panel discussions about privacy, and the NTIA stakeholder meetings were very positive, working towards best practices and guidelines within the commercial use cases. Indeed, the best practices and use cases that came out of the NTIA meetings were very strong. I wish we had more input from the privacy sector, because there is space to meet in the middle and bring face recognition forward. A productive discussion would be to elaborate on privacy issues within a specific, defined use case.
Everything is not negative – face recognition can be a powerful tool in terms of identifying a person. It does have its limitations, but in the right use cases it is a force multiplier. Of course, the perception of being invasive or creepy is something that the industry needs to work on and we should not be afraid to engage – private, public figures and even advocates. It isn’t helpful to just deny or distort realities.
CG: Certainly makes a lot of sense, shifting on the business side of face recognition, what vertical markets are seeing pronounced growth at the moment?
TW: Firstly, three markets come to mind, and I know these are often related, but we have seen demand in border security, public security and travel – the latter also including hospitality and tourism.
We are all tired of the long, long lines in airports at peak tourism season and technology can help solve this. This is another area where facial recognition can be a force multiplier. It can enhance the travel experience, making it more seamless, and that will lead into the tourism industry.
Another sector that’s seeing strong growth is public security, for instance in law enforcement as an enhanced investigation tool. We can see a number of agencies using mugshots to identify suspects, victims and in all sorts of ways. The increase in video data that we have to analyse events will also see facial recognition play a larger role in security.
Rob Sprecher: I certainly agree, and I think that with the quality of video that is becoming available, that there are going to be incremental expectation changes over how video surveillance is used. For example, researchers are developing methods that can assess an individual’s attitude, such as whether a person is angry or anxious.
Factors like that or looking to a customer service viewpoint on big data and analytics, are actually just scratching the surface of the changes and improvements we are seeing. We are no longer just searching for faces, but finding faces in the context of an interaction. These advances will be very important in the use of video analytics and big data.
CG: Thanks Rob, can you relate some of the innovations you are working on for face and analytics?
RS: Craig, the things that we are working now include the integration of face recognition into multi-modal solutions. Importantly, we are leveraging the accuracy of facial recognition for law enforcement and border officials who need rapid identification of individuals out in the field.
The technology gives these officials a chance to very quickly establish identities, using a tool like our Morpho Face Detective. Rather than having to rely on an expert back in a lab, they can create investigative leads in the field while being close to the action.
What we are finding is that as the tech improves, we are seeing incremental rise in expectations around finding faces, and these expectations can drive innovations. When people ask questions over aspects such as performance in crowds, or on toll roads, for example, then we know what areas to work on. I think wider adoption of face recognition will generate a new range of applications as people consider its potential.
CG: Why is face becoming so prevalent in smart border solutions?
RS: Because it is a frictionless solution. It doesn’t require an individual to stop and place their fingers on a fingerprint scanner or look intently into an iris lens. The quality of face detection and recognition has improved to the point that zone of capture is much larger and faster.
TW: The biometric industry has made significant strides in improving the usability, reliability and accuracy of facial recognition. Not only has facial recognition become a viable solution, it has become the cornerstone of secured yet expedited travel on a large scale.