On Facial Age Estimation, improvements and updates

profile picture Matt Prendergast 3 min read
People with their estimated ages

We continue to invest in improving the accuracy of our world-leading facial age estimation and we will soon be releasing a new ‘Jan 2023’ model with improved accuracy across all skin tones and gender for those aged 6-70.

At Yoti we constantly strive to improve all our services given our commitment to tech for good, and given the growing market importance of these services and increasing regulatory engagement. For example, coming soon we’ll be introducing multiple, concurrent age estimation models, which will help improve accuracy even further, whilst still completing a check within seconds. 

We will also continue to update our Yoti Age Estimation white paper, now 4 years old this month, in which we transparently publish our accuracy across age, gender and skin tones by year between ages 6 and 70. We expect to publish the updated white paper in the next couple of weeks but importantly, we can share today our Jan 2023 model further improves our accuracy and further reduces bias.

Across all age ranges, gender and skin tones we are seeing a 5.1% increase in accuracy, reducing the weighted average MAE for 6 to 70 year olds from 3.0 to 2.8 years. As you can see below for this model we have been focussing on reducing the discrepancy between light skintones (tone 1) and darker skin tones (tone 3) We recognise our work is not finished in this regard.

Table displaying yoti's facial age estimation improvements over time, with an average of 5.1% improvement across all skin tones and genders from May 22 to Jan 23
Table shows mean average error (MAE) for age range 6-70 by gender and skin tone, and the percentage improvement from our May 2022 model to our January 2023 model.

So, well done for marking your own homework, you might say! But we are happy to report our May 2022 white paper has been independently verified as to the measurement methodology and accuracy of our results. On the request of one of our clients, the ACCS undertook an independent evaluation of our May 2022 white paper and had this to say:

“The training, testing and results reporting presented in the whitepaper have been independently validated by ACCS, who have certified that Yoti have deployed appropriate methodologies to analyse the performance of their Age Estimation algorithm, including ensuring appropriate separation of machine learning training data, testing data and validation data.”

In addition to this, we are also happy to announce that we have been invited to participate in a workshop on ICO commissioned research on the measurement of age assurance on Thursday 19th January, in London organised by ACCS and AVPA. 

Follow us on LinkedIn to stay abreast of Yoti news or get in touch to find out more. 

Related stories

An image showing three icons, which represent face matching, multi-factor verification and document authenticity checks. These are three of the methods discussed in the article.

Strengthening existing age verification methods to meet modern challenges

In today’s digital landscape, effective age verification is key to helping us build a safer, more trusted online world. Whether it’s ensuring young people have age-appropriate experiences or preventing underage sales of age-restricted items, accurate age checks are crucial for safeguarding online users. We’re seeing a wave of age-related legislation come in around the world. With this, online services face the significant challenge of balancing user privacy with user safety. Alongside newer age-checking solutions, existing age verification methods still have their benefits. We look at how these methods can be strengthened for greater security, efficiency and user privacy.  

7 min read
An image showing a comparison between two age results on a mobile phone screen. On the left hand side is a screen showing an age result that has been verified with Yoti's facial age estimation. The right hand side shows a screen with a showable Digital ID card that has been verified with a UK driving licence. Both screens show an "18+" age result.

Age estimations in our Digital ID app

Digital ID users can have their age estimated in the app, and then anonymously share the age result. This gives them a secure, private and easy way to pass age checks. – With lots of legislation being introduced to improve online safety, including the UK Online Safety Act and the EU Digital Services Act, more businesses are looking at how to verify the age of users. People should have a choice in how they prove their age, so they can choose the method that works best for them. They might use their identity document, complete a facial age estimation

5 min read
An image of someone using their laptop. The screen shows the person choosing which method they'd like to use to prove their age.

Introducing Yoti Keys: privacy-focused, seamless and anonymous age verification

We’re excited to announce a new age assurance solution: Yoti Keys.  Yoti Keys let people verify their age once and gain continued access to an ecosystem of websites without having to prove their age again, regardless of if they are using an incognito or private browser. A Yoti Key, using passkey technology, doesn’t store any personal information. This helps people to remain completely anonymous but verified – all on their device.   How a Yoti Key works Complete an age check: The user completes an initial age check to prove their age. They can choose which method to use,

4 min read