Self-supervised visual learning for analyzing firearms trafficking activities on the Web
This addresses a critical real-world problem for public security and law enforcement by identifying criminal networks, but it is incremental as it highlights an unexplored area without presenting new results.
The paper tackled the problem of automated visual firearms classification from RGB images for analyzing firearms trafficking on the web, and found that neither Visual Transformer architectures nor Self-Supervised Learning approaches have been evaluated on this task, indicating a gap in current methods.
Automated visual firearms classification from RGB images is an important real-world task with applications in public space security, intelligence gathering and law enforcement investigations. When applied to images massively crawled from the World Wide Web (including social media and dark Web sites), it can serve as an important component of systems that attempt to identify criminal firearms trafficking networks, by analyzing Big Data from open-source intelligence. Deep Neural Networks (DNN) are the state-of-the-art methodology for achieving this, with Convolutional Neural Networks (CNN) being typically employed. The common transfer learning approach consists of pretraining on a large-scale, generic annotated dataset for whole-image classification, such as ImageNet-1k, and then finetuning the DNN on a smaller, annotated, task-specific, downstream dataset for visual firearms classification. Neither Visual Transformer (ViT) neural architectures nor Self-Supervised Learning (SSL) approaches have been so far evaluated on this critical task..