About the Project
The investigation of child sexual abuse material (CSAM) has traditionally relied on a process of matching seized media against databases of known hash values and, more recently, the deployment of facial recognition technologies. However, at present, these matching techniques are limited in their utility, as they identify that the media is CSAM but provide no additional information, such as whether it is of a known victim. Thus, they require considerable manual intervention by investigators.
This research project seeks to improve current investigative practices via the development of a software tool that is specifically designed to ingest media files containing CSAM, extract multiple biometric attributes, and match subjects across videos based on these biometric attributes. This software is designed to not only enhance existing analytical capabilities, by addressing some of the shortcomings flagged above, but also augment other automated CSAM detection and collection technologies. To this end, our software is designed to be interoperable with web crawlers, including our own (see additional resources, below), to ‘crawl’ the Surface and/or Deep Web to detect and gather CSAM for subsequent analysis.
This research project is being undertaken in partnership with the Australian Institute of Criminology, and involves a multinational, and interdisciplinary team of criminologists, computer scientists, psychologists, and engineers.
Watch our latest research update: February 2022
Adelaide Lab researchers and contacts
Associate Professor Russell Brewer
Dr Katie Logos
Mr Thomas Swearingen
Research Partners
Associate Professor Bryce Westlake, San Jose State University
Professor Arun Ross, Michigan State University
Dr Dana Michalski, Defence Science and Technology Group