Countering Online Child Exploitation
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.
Detecting Illegal Firearms
We are using state-of-the-art 3D-scanning and photogrammetry technology to develop a bespoke hybrid dataset of real-world firearm imagery. This will, in turn, drive the development of high-quality and granular machine learning models for the classification of firearms.
Managing Sensitive Data
One of our major research themes is the design of infrastructure for facilitating ‘eyes-off’ access to restricted data in the pursuit of collaborative data science research and development. Originally conceived for the development of classification models for CSAM in conjunction with law enforcement, we are discovering applications for this approach across a wide variety of similar domains.
Curating Ethical Datasets
We are going back to first principles, creating and managing datasets of benign ‘in-the-wild’ material for research use with the knowledge and consent of those depicted. In particular we are concerned with how the data is captured in the first place, and how the consent for its use is managed through time.
Interdisciplinary AI Research
We are working on AI in judicial contexts where explainability of algorithms is crucial. We are also working on the technical, criminological and legal responses to deepfake technologies given their potential serious harms via image-based abuse.
