Facial recognition and the new challenge of “stay at home” learning and examinations

9/4/20, 8:53 am

March 2020 will go down as the month when everything in our lives changed, including the way our children were educated.

Although the School of the Air, an Australian innovation, has provided education through internet technologies since 2009 to students on remote cattle and sheep stations, mainstream education in Australia has always relied on traditional or classroom learning. From chalkboards to whiteboards, “bums on seats” in class or lecture rooms has been the accepted norm.

For some time the pendulum has been shifting. Now, by necessity, State and Territory leaders have been asking parents to have their children schooled from home while coronavirus cases increase across Australia. In many jurisdictions, that has moved to mandatory schooling from home.

COVID-19 will create a whole new frontier in remote learning, a forced frontier.

The new challenges for teaching institutions such as schools, universities and vocational training providers are many – not just in terms of delivering content but also tracking student attendance and verifying the student’s identity. The fact is, in tertiary and vocational training environments, a teacher or training instructor is less likely to have a one-on-one relationship with the students in a way that say a high school teacher would.

This is particularly crucial when it comes to examinations. Statistical evidence seems to indicate that attempts at exam fraud through student impersonation is a real problem, even here in Australia.

Starting with Year 12 exams, the stakes get progressively higher and the integrity and scrutiny of the examination delivery framework needs to be at its highest. Students and parents are, rightly or wrongly, of the opinion that careers and futures are on the line and that the people they sit the exam with are indeed the people they should be competing against, whether in the same school or institution, or another.

If, and it remains a big IF at the moment, exams or assessments have to be conducted online in an era of social distancing, then verification, proof through facial recognition and remote invigilation of strictly controlled out of exam room situations, may end up becoming a reality.

Facial recognition - an answer to verification

But is eyeballing a student in a remote classroom to check them off the attendance roll or authenticate them in a remote exam environment even a possibility? NEC Australia and its solution partners will tell you that both are possible using facial recognition.

Facial recognition, one of the world’s most rapidly changing areas of biometrics, can authenticate, with an incredibly high degree of accuracy, the identity of an individual in a classroom or in an examination situation.

NEC and its Melbourne-based partner Genix Ventures, have already engineered and launched NEC ID, a biometrics solution that ensures that a person who has enrolled for a class or a course, is in fact the same person taking the course or more importantly, sitting the exam, and not an impersonator.

COVID-19 is about to force educational institutions to investigate ways of establishing a student’s identity. It is a challenge, but one that has an answer – facial recognition. Through its accuracy, facial recognition can do what even an invigilator in an exam room situation cannot – it can establish a student’s legitimacy via algorithms that are becoming more advanced and accurate.

Facial recognition works by calculating and assessing the relationships between facial landmarks on an image and then by comparing these measurements to other facial images. This is pattern matching. NEC’s facial recognition technology utilises artificial intelligence to create a 3D model of a given face, correcting its position so it can be compared to another facial image more accurately in non-standardised photos and difficult non-cooperative environments (known as "on the move" capture).

Through its Australia-designed platform NEC ID, images collected by NEC Australia are immediately converted into a hashed template by its industry-leading algorithm. This template is then stored by the engine for future use against other live images. The data in these templates can only be used with the biometrics engine hosted in NEC Australia’s highly secure datacentre, meaning it cannot be used by third parties.

NEC has again been ranked number one globally for its facial recognition algorithm amongst entries from companies in almost 50 countries by the US National Institute of Standards and Technology (NIST).

If your learning institution wishes to learn more about NEC’s facial recognition solutions and its applications to assist you, see our areas of expertise.

Roger White
Corporate Communications Manager