AiLECS Lab to create ethically-sourced face biometric database to fight child abuse
The Australian Federal Police (AFP) and Monash University entered a new collaboration aimed at creating an ethically-sourced face biometrics database to combat child exploitation.
The project is a brainchild of the AiLECS Lab (AI for Law Enforcement and Community Safety), an entity designed to bring together researchers from Monash University’s Faculty of Information Technology and AFP, with a particular focus on artificial intelligence (AI)-based projects for public safety.
Dubbed ‘My Pictures Matter,’ the crowdsourcing campaign is asking individuals aged 18 and above to submit photographs of themselves as children.
The files will then be utilized to train biometric AI models that will potentially “recognize the presence of children in ‘safe’ situations, to help identify ‘unsafe’ situations and potentially flag child exploitation material.”
The creation of the database was prompted by the fact that often, AiLECS Lab Co-Director Associate Professor Campbell Wilson explains, similar machine learning models are trained with images that are scraped off the internet or without documented consent for their use.
“Sourcing these images from the internet is problematic when there is no way of knowing if the children in those pictures have actually consented for their photographs to be uploaded or used for research,” Wilson adds.
“By obtaining photographs from adults, through informed consent, we are trying to build technologies that are ethically accountable and transparent.”
For context, the AFP admitted using Clearview’s facial recognition technology in 2020. Since then, the company has been involved in a series of court cases for allegedly scraping users’ face biometrics data from social media.
Most recently, Clearview has been fined approximately £7.5 million (US$9.5 million) by the UK’s Information Commissioner’s Office (ICO).
To avoid scenarios like this, and to enable user control over their privacy, all data collected as part of the My Pictures Matter is going to be stored and managed according to the principles of data minimization.
“We are not collecting any personal information from contributors other than the email addresses associated with consent for research use, and these email IDs will be stored separate to the images,” explains Project lead, Dr Nina Lewis.
According to these guidelines, contributors will be able to get details and updates about each stage of the research, as well as opt to change user permissions or revoke their research consent and withdraw images from the database at a later date.
The project aims to build a database of 100,000 ethically-sourced images by the end of 2022.
Yoti began building a database of images of children submitted by their parents to train its age estimation technology in 2021.