News
Plug to Place: Indoor Multimedia Geolocation from Electrical Sockets for Digital Investigation
Congratulations to Kanwal Aftab and co-authors Graham Adams, and Mark Scanlon on the publication of Plug to Place: Indoor Multimedia Geolocation from Electrical Sockets for Digital Investigation in Forensic Science International: Digital Investigation.
Co-authors: Graham Adams, and Mark Scanlon.
AI-generated summary of the contribution: This paper introduces a pipeline that uses electrical sockets as consistent indoor markers for geolocation, addressing the challenges of similar room layouts, frequent renovations, visual ambiguity, and limited datasets in sensitive domains. The three-stage deep learning pipeline detects plug sockets, classifies them into one of 12 plug socket types, and maps the detected socket types to countries. The approach is evaluated on the Hotels-50K dataset, focusing on the TraffickCam subset of crowd-sourced hotel images, and demonstrates its practical utility for law enforcement in human trafficking investigations. The paper also presents two dedicated datasets: a socket detection dataset of 2,328 annotated images expanded to 4,072 through augmentation, and a classification dataset of 3,187 images across 12 plug socket classes. The code, trained models, and data for this paper are available open source.