Intelligent traffic systems for railway object detection have become the focus of research in recent years. Accurate and fast object detection using a camera is an important but challenging problem in the railway industry. In this study, we propose an object detector with low power yet robust detection for collision warning in a train safety system. The proposed object detector comprises three modules. First, a stable sampling module is used to reduce the dimensions of the feature map and image information loss. Second, a lightweight feature extraction module utilizes a dynamic bottleneck structure to control the calculation load and enhance the expressive ability of the model. Third, a feature-fusion module combines high- and low-level features to enhance the semantic information and improve the accuracy of detection of multiscale and small objects. Experimental results demonstrate that the proposed network achieves reasonable results for railway object detection and outperforms the current state-of-the-art detectors. Finally, we design an obstacle-avoidance device that can be installed at the front of the train for real-time security warning in real-world conditions.

Detecting time-series anomalies is extremely intricate due to the rarity of anomalies and imbalanced sample categories, which often result in costly and challenging anomaly labeling. Most of the existing approaches largely depend on assumptions of normality, overlooking labeled abnormal samples. While anomaly assumptions based methods can incorporate prior knowledge of anomalies for data augmentation in training classifiers, the adopted random or coarse-grained augmentation approaches solely focus on pointwise anomalies and lack cutting-edge domain knowledge, making them less likely to achieve better performance. This paper introduces CutAddPaste, a novel anomaly assumption-based approach for detecting time-series anomalies. It primarily employs a data augmentation strategy to generate pseudo anomalies, by exploiting prior knowledge of anomalies as much as possible. At the core of CutAddPaste is cutting patches from random positions in temporal subsequence samples, adding linear trend terms, and pasting them into other samples, so that it can well approximate a variety of anomalies, including point and pattern anomalies. Experiments on standard benchmark datasets demonstrate that our method outperforms the state-of-the-art approaches.

Mobile cloud computing has been widely used to offload the execution of computation-intensive applications to a cloud for mobile devices. One concern of computation offloading is the energy consumption of transmission due to the long air interface distance between the user and the cloud. Therefore, we introduce device-to-device communications into computation offloading, and then propose a framework of user-centric local mobile cloud-assisted D2D communications in heterogeneous cloud RANs (H-CRANs). In this framework, resource-poor devices can utilize other users' vacant computing resources. Following that, we consider an energy- aware subtask allocation strategy with limited fronthaul data transfer in H-CRAN. Finally, simulation results illustrate that the proposed strategy consumes less transmission energy than conventional computation offloading strategies with fronthaul usage fairness guaranteed.

Ultrasonography has become an essential part of clinical diagnosis owing to its noninvasive, and real-time nature. To assist diagnosis, automatically segmenting a region of interest (ROI) in ultrasound images is becoming a vital part of computer-aided diagnosis (CAD). However, segmenting ROIs on medical images with relatively low contrast is a challenging task. To better achieve medical ROI segmentation, we propose an efficient module denoted as multiscale attentional convolution (MSAC), utilizing cascaded convolutions and a self-attention approach to concatenate features from various receptive field scales. Then, MSAC-Unet is constructed based on Unet, employing MSAC instead of the standard convolution in each encoder and decoder for segmentation. In this study, two representative types of ultrasound images, one of the thyroid nodules and the other of the brachial plexus nerves, were used to assess the effectiveness of the proposed approach. The best segmentation results from MSAC-Unet were achieved on two thyroid nodule datasets (TND-PUH3 and DDTI) and a brachial plexus nerve dataset (NSD) with Dice coefficients of 0.822, 0.792, and 0.746, respectively. The analysis of segmentation results shows that our MSAC-Unet greatly improves the segmentation accuracy with more reliable ROI edges and boundaries, decreasing the number of erroneously segmented ROIs in ultrasound images.