PAD-F: Prior-Aware Debiasing Framework for Long-Tailed X-ray Prohibited Item Detection

TL;DR

This paper introduces PAD-F, a novel framework for long-tailed prohibited item detection in X-ray images, combining material-aware data augmentation and co-occurrence feature aggregation to improve detection of rare classes.

Contribution

The paper proposes a new debiasing framework that leverages material and co-occurrence priors, with novel augmentation and feature modules, for improved long-tailed object detection in X-ray security images.

Findings

Achieves up to +17.2% AP50 improvement for tail classes.

Significantly outperforms existing state-of-the-art methods.

Effective across multiple popular detectors.

Abstract

Detecting prohibited items in X-ray security imagery is a challenging yet crucial task. With the rapid advancement of deep learning, object detection algorithms have been widely applied in this area. However, the distribution of object classes in real-world prohibited item detection scenarios often exhibits a distinct long-tailed distribution. Due to the unique principles of X-ray imaging, conventional methods for long-tailed object detection are often ineffective in this domain. To tackle these challenges, we introduce the Prior-Aware Debiasing Framework (PAD-F), a novel approach that employs a two-pronged strategy leveraging both material and co-occurrence priors. At the data level, our Explicit Material-Aware Augmentation (EMAA) component generates numerous challenging training samples for tail classes. It achieves this through a placement strategy guided by material-specific absorption rates and a gradient-based Poisson blending technique. At the feature level, the Implicit Co-occurrence Aggregator (ICA) acts as a plug-in module that enhances features for ambiguous objects by implicitly learning and aggregating statistical co-occurrence relationships within the image. Extensive experiments on the HiXray and PIDray datasets demonstrate that PAD-F significantly boosts the performance of multiple popular detectors. It achieves an absolute improvement of up to +17.2% in AP50 for tail classes and comprehensively outperforms existing state-of-the-art methods. Our work provides an effective and versatile solution to the critical problem of long-tailed detection in X-ray security.