A New GAN-Based Anomaly Detection (GBAD) Approach for Multi-Threat Object Classification on Large-Scale X-Ray Security Images

Abstract

Threat object recognition in x-ray security images is one of the important practical applications of computer vision. However, research in this field has been limited by the lack of available dataset that would mirror the practical setting for such applications. In this paper, we present a novel GAN-based anomaly detection (GBAD) approach as a solution to the extreme class-imbalance problem in multi-label classification. This method helps in suppressing the surge in false positives induced by training a CNN on a non-practical dataset. We evaluate our method on a large-scale x-ray image database to closely emulate practical scenarios in port security inspection systems. Experiments demonstrate improvement against the existing algorithm.

End-to-End Object Separation for Threat Detection in Large-Scale X-Ray Security Images

“…Ultimately, the end-to-end object separation framework benefits by incorporating all the proposed updates. Method TPR FPR [9] 79.50 0.40 [13] 92.80 4.00 [14] 94.40 7.70 [5] 95.57 2.17 [2] 97.33 1.46 Proposed 96.87 1.24…”

Advanced security testing using a cyber‐attack forecasting model: A case study of financial institutions

“…Fortinet ® is a one of the pioneer service providers in Jordan and the world for the multi-threat security solutions of the next generation. 32 This increased the reliability of the generated data. Both WAF and IDS/IPS are network security applications that are used for network and data protection.…”

“…The log registers of network attacks targeting one of the Jordanian banks were generated from two network security systems: Fortinet® WAF and Fortinet® FortiAnalyser IDS/IPS. Fortinet® is a one of the pioneer service providers in Jordan and the world for the multi‐threat security solutions of the next generation 32 . This increased the reliability of the generated data.…”

Multiscale generative model using regularized skip-connections and perceptual loss for anomaly detection in toxicologic histopathology

“…For example, it is critical to identify potential weapons manifesting as common, hidden, or unfamiliar “abnormalities” in normal X-ray security passenger screening data. 5 , 6 Similarly, industrial production is rich with normal data, but detecting random anomalies is crucial for quality assurance. 7 , 8 Developments of anomaly detection in the medical domain include studies on chest X-ray, 9 , 10 brain MRI, 11 , 12 , 13 and retinal OCT 14 , 15 , 16 to screen for potential signs of disease.…”