Countering online child exploitation
The Australian Centre to Counter Child Exploitation received in 2020 alone, over 21,000 reports of the online child sexual exploitation. Each of these reports may contain large volumes of images and videos of children being abused in horrific ways.
The US based National Center for Missing & Exploited Children CyberTipline has since 1998 reviewed over 322 million images and videos of suspected child sexual exploitation.
In addition to the horrendous life-long impacts on the children targeted, their abuse has a ripple effect on society, deeply affecting all those involved in responding to, investigating and prosecuting this crime.
We are developing AI techniques for automated detection and triage of child sexual exploitation material. This is a particularly difficult task, not least because of the sensitivities involved in dealing with training data these algorithms use to learn to recognise this material.
AiLECS work in this area currently includes:
- transfer learning approaches to boosting the performance of image and video classification algorithms that detect child exploitation material,
- examining the robustness of perceptual hashing algorithms, that can be used to rapidly identify previously seen images or videos,
- the building of infrastructure that allows for safer training of future machine learning algorithms for detecting child exploitation material,
- curating training datasets in an ethical and consentful manner through our VALID project.