Search

During X-ray inspection detection, a detector converts the collected X-rays from objects into electrical signals, which are then transmitted to a computer for image processing and analysis. From the aspect of digital image processing, detection tasks mainly focus on data processing and transformation to identify valuable features, which make the algorithms more effective. The consistent requirement for speed and accuracy in X-ray prohibited item detection is still not fully satisfied, especially in pictures obtained under special imaging conditions. For noisy X-ray images with heavy occlusion, a direct and suitable approach of representation learning is the optimal solution. According to our study, we realized that heterogeneous information fusion from different extraction approaches can be applied effectively to overcome this issue. We proposed two innovative algorithms to extract effective features of X-ray objects to significantly improve the efficiency of X-ray prohibited item detection. The brief model we proposed fuses the representations learned from the noisy X-ray images and outperforms the best model (DOAM-O) so far on OPIXray. Furthermore, the attention module we designed to select information on deep learning and representation strengthens the model; considering this, the model utilizes lesser time for both training and inference, which makes it easier to be trained on a lightweight computing device.

In the realm of visual tracking, remote sensing videos captured by Unmanned Aerial Vehicles (UAVs) have seen significant advancements with wide applications. However, there remain challenges to conventional Transformer-based trackers in balancing tracking accuracy and inference speed. This problem is further exacerbated when Transformers are extensively implemented at larger model scales. To address this challenge, we present a fast and efficient UAV tracking framework, denoted as SiamPT, aiming to reduce the number of Transformer layers without losing the discriminative ability of the model. To realize it, we transfer the conventional prompting theories in multi-model tracking into UAV tracking, where a novel self-prompting method is proposed by utilizing the target’s inherent characteristics in the search branch to discriminate targets from the background. Specifically, a self-distribution strategy is introduced to capture feature-level relationships, which segment tokens into distinct smaller patches. Subsequently, salient tokens within the full attention map are identified as foreground targets, enabling the fusion of local region information. These fused tokens serve as prompters to enhance the identification of distractors, thereby avoiding the demand for model expansion. SiamPT has demonstrated impressive results on the UAV123 benchmark, achieving success and precision rates of 0.694 and 0.890 respectively, while maintaining an inference speed of 91.0 FPS.