Search by product, support information & more
Most popular searches:
Search for "PA00185"Read our Latest Blogs >
Click here for Technical Support >
We’ve got a new look! Please note: all users need to re-register to take advantage of our new website functionality
|
Artificial Intelligence (A.I.) applied in Biometrics is being widely used in computer science as a form of identification and access control. Biometrics comes in many forms such as fingerprints, facial recognition, iris scanners and is already being used by many organisations, apps and services. It is a way to identify human and behavioural characteristics and although it may have created some controversy in the past, consumers are becoming more relaxed with the technology as it becomes frequently used in everyday life. The younger generation especially are more comfortable with biometrics as shown in a recent IBM report which found that 75% of millennials were happy to use biometrics compared with just 58% of those over 55 (Source: www.capitaidentitysolutions.co.uk). The use of biometric technology in the retail sector is increasingly being used to improve the customer experience and efficiencies, and to differentiate retailers in this fiercely competitive environment (Source: www.bayometric.com). The self-checkout market is undoubtedly on the up – RBR a London based research organisation, predicted that the global installed base of self-checkout terminals will rise from 191,000 in 2013 to reach nearly 325,000 by 2019 (Source: (www.technologyrecord.com), therefore the uptake of biometric technology is certainly set to rise.
Biometric technology can also be used to grant access to age-restricted premises such as Adult Gaming Centres or age restricted gaming machines. This could help remove human error from current exclusion systems (Source: www.biometricupdate.com). Age recognition software could also be a way to tackle the growing problem of underage gamblers. Indeed, officials in Japan are proposing the installation of facial recognition systems in pachinko parlours and other gambling premises. Using age recognition solutions and technology to help gaming venues refuse entry to those underage or excluded could help combat this growing social issue (Source: www.gamblinginsider.com).
ICU from Innovative Technology is a brand-new device that uses intelligent biometric algorithms to automate age verification for controlling access to age restricted purchases and premises. Simple integration makes the ICU suitable for numerous applications including retail self-serve, kiosks and gaming. Automation is key; especially for retail self-service check-outs where both customers and retailers can benefit. ICU boasts a 96% accuracy rate (detecting <18s) compared to a human accuracy rate of 69%* (Source: https://innovative-technology.com/images/pdocuments/datasheets/ICU_TechnicalPaper.pdf) leading to faster throughput and increased revenue. As well as having market leading accuracy ICU has anti-spoof detection in the form of its 3D anti-fraud system, which detects 2D photographs in print or digital form. It is a standalone device, eliminating the need for expensive hardware, third party fees and ongoing costs, offering the lowest cost of ownership of any age verification solution. Installation is easy as ICU can be mounted horizontally or vertically via custom bracket, standard camera mount or inside the machine via a through panel.
The ICU utilises machine learning to verify the age of the subject. This is an area of software development where the underlying algorithm learns how a face of a certain age will generally appear. This is achieved by training the algorithm with several thousand faces of known ages. It extracts specific features and associates those features with the specific known ages – so for example the algorithm learns that a subject under 18 will have a certain cohort of feature definitions compared to that of a subject over 25. When a subject presents to the ICU, it extracts those facial features and from what it has learnt, it returns the most likely age of the subject. In a similar manner to how humans perceive age, the computer uses key identifiers to estimate age but crucially is not swayed by other human factors such as tiredness, emotion or subconscious bias.
A paper by The Royal Society Open Science (Two sources of bias explain errors in facial age estimation by Clifford, Watson and White) found that error in human guesses across an age range of 7 to 70 approaches ±8 years. There are two sources of bias to explain errors in facial age estimation. First, estimates tend to be biased towards the age of the preceding face. Second, humans tend to systematically underestimate the ages of older people, and over-estimate the age of younger people (Source: https://royalsocietypublishing.org/doi/10.1098/rsos.180841)
Despite the massive rise in self-service checkouts, the RBR study found that 93% expressed frustration with them and age restrictions was cited as one of the major factors behind this frustration. Similarly, in a poll undertaken by Newton Consultancy, 48% of consumers rate slow checkouts as most frustrating element of shopping (Source: www.thegrocer.co.uk). Many times, we have waited for (human) authorisation at the checkout for age restricted products, whereas today with ICU, we can know in a matter of seconds whether a person needs to present ID or not. For the customer, this brings a better experience, for the operator, more income & lower costs, and for the integrator, a considerable product differentiator. Currently, the device will flag when a secondary form of age verification is necessary (when the device deems the subject to be under 25 for example). This secondary verification would be in the form of a physical ID. A reasonable solution would be to combine human checks with ICU Automated Age Verification technology.
ICU does not store any images or personal data so there are no implications for GDPR and no privacy concerns for the operator. We have acknowledged that age restrictions for different products and premises vary from country to country so this global product can be easily tailored to meet the age restrictions and specific regulations of any country. For spoof detection ICU uses multiple cameras to detect if the subject is ‘3D’. A photograph of a person is 2D so the unit can detect that easily. Further to this it can also detect if a photograph is curved (to make it a pseudo 3D object). The fact that ICU can detect curved objects is a great differentiator and one of its key features.
High level of accuracy: A technical study undertaken by ITL to test the accuracy of the ICU found that the device can achieve an accuracy of 96% in detecting subjects under the age of 18
Its quick – complete process takes less than 5 seconds: With no need to wait for approval, ICU can speed up the process and allows for greater throughput, quicker transactions, shorter queues and as a result an enhanced experience for the customer
Lowest costs of ownership
Combats fraud: Uses an 3D anti-fraud system for anti-spoof detection
Outperforms humans: A further test demonstrated that ICU also out-performed humans (accuracy rate 69%) in age estimation with no fatigue or human error affecting the results. Plus, it also takes away the emotion of identifying people.
Simple integration: Standalone, complete unit so integration is simple
Ease of use: Users do not need to pre-register
For more information and prices please do not hesitate to contact me: aobrien@innovative-technology.com
Sources:
https://www.bayometric.com/use-of-biometric-technology-in-the-retail-sector/
https://www.biometricupdate.com/201803/japan-considering-biometric-identification-for-casino-access
https://www.gamblinginsider.com/news/6891/japan-wants-facial-recognition-at-gambling-establishments
https://royalsocietypublishing.org/doi/10.1098/rsos.180841:
*Based on Innovative Technology investigation: https://innovative-technology.com/images/pdocuments/datasheets/ICU_TechnicalPaper.pdf