News
Perceptual Colour-based Geolocation of Human Trafficking Images for Digital Forensic Investigation
Congratulations to Jessica Herrmann and co-authors Opeyemi Bamigbade, John Sheppard, and Mark Scanlon on the publication of Perceptual Colour-based Geolocation of Human Trafficking Images for Digital Forensic Investigation in 2024 Cyber Research Conference - Ireland (Cyber-RCI).
Co-authors: Opeyemi Bamigbade, John Sheppard, and Mark Scanlon.
AI-generated summary of the contribution: This research explores the role of colour in Content-Based Image Retrieval (CBIR) for human trafficking image analysis. The study evaluates the effectiveness of colour-based descriptors in identifying hotel rooms in a large dataset of images from different sources. The research examines the optimal number of colour values for maximising matching accuracy and precision, with the aim of assessing the procedure’s performance and providing insights for improving image-based hotel identification. The study achieves a Top-50 accuracy of over 95% on the Hotels-50K dataset, demonstrating the potential of colour-based descriptors in advancing image analysis tools for human trafficking investigations and other contexts.