SPS Webinar: 24 February 2022, by Dr. Samet Akcay - Recent Advances of Deep Learning within X-ray Security Imaging

Upcoming SPS Webinar!

Title: Recent Advances of Deep Learning within X-ray Security Imaging
Date: 24 February 2022
Time: 11:00 AM EST (New York time)
Duration: Approximately 1 Hour
Presenters: Dr. Samet Akcay

Based on the IEEE Xplore® article: Using Deep Convolutional Neural Network Architectures for Object Classification and Detection Within X-Ray Baggage Security Imagery
Published: IEEE Transactions on Information Forensics and Security, March 2018
Download: Original article will be made freely available for download for 48 hours from the day of the webinar, on IEEE Xplore®

Abstract:

X-ray security screening is widely utilized in aviation and transportation, and its importance has sparked interest in automated screening systems. The goal of this webinar is to explore computerized X-ray security imaging methods by classifying them into traditional machine learning and modern deep learning applications. The talk briefly reviews the traditional machine learning methodologies used in X-ray security imaging, and subsequently delves deeper into the applications of recent deep learning-based algorithms. The suggested taxonomy divides deep learning applications into supervised and unsupervised learning categories, with a focus on object categorization, detection, segmentation, and anomaly detection. The session goes on to look at some well-known X-ray datasets and presents a performance benchmark. The talk will be concluded with a discussion and future directions for X-ray security imagery, based on current and future advances in deep learning.

Biography:

Dr. Samet Akcay received his PhD degree in the Department of Computer Science at Durham University, UK. Prior to pursuing a PhD, he received his MSc degree from the Robust Machine Intelligence Lab at the Department of Electrical Engineering at Penn State University, USA.

He is currently an AI Research Engineer/Scientist at Intel. He is working on self-supervised anomaly detection and localization for industrial and medical applications and recently open-sourced one of the largest anomaly detection libraries in the field. His primary research interests are real-time image classification, detection, anomaly detection, and unsupervised feature learning via deep/machine learning algorithms.

Dr. Akcay has published numerous academic papers in the leading computer vision and machine/deep learning conferences and journals and contributed to several projects funded by the UK Home Office, Department for Transport, and the Ministry of Defence.