X-ray inspection of pipeline welds brings benefits for the industrial quality and safety of production. However, noise and complex texture structure in X-ray images caused by issues such as poor contrast and blurry defects lead to extraordinary obstacles for lightweight and high-precision identification to satisfy almost 100% accuracy requirements, especially in constrained environments. This paper proposes a novel method, named YOLO-Xweld, to detect pipeline welding defects for constrained application scenarios. YOLO-Xweld is derived from YOLO v3-tiny by adding an SPP module to the backbone network to enhance multi-scale feature extraction, and weakening large-scale network detection branch based on characteristics of pipeline weld X-ray images. Depthwise separable convolution is introduced to achieve a lightweight model and reduce the dependence on the high-performance hardware. Experiments show that the number of the model's parameters and floating-point operations have dropped by over 70%, and the model size has been reduced to 1/15 of the original size, while the test accuracy is close to 100%. This is of great significance in reducing the dependence on high-performance hardware and enhancing the convenience and operability during inspections in the constrained pipeline weld detection environment.

Scene graph generation (SGG) aims to detect objects in an image along with their pairwise relationships. There are three key properties of scene graph that have been underexplored in recent works: namely, the edge direction information, the difference in priority between nodes, and the long-tailed distribution of relationships. Accordingly, in this paper, we propose a Graph Property Sensing Network (GPS-Net) that fully explores these three properties for SGG. First, we propose a novel message passing module that augments the node feature with node-specific contextual information and encodes the edge direction information via a tri-linear model. Second, we introduce a node priority sensitive loss to reflect the difference in priority between nodes during training. This is achieved by designing a mapping function that adjusts the focusing parameter in the focal loss. Third, since the frequency of relationships is affected by the long-tailed distribution problem, we mitigate this issue by first softening the distribution and then enabling it to be adjusted for each subject-object pair according to their visual appearance. Systematic experiments demonstrate the effectiveness of the proposed techniques. Moreover, GPS-Net achieves state-of-the-art performance on three popular databases: VG, OI, and VRD by significant gains under various settings and metrics. The code and models are available at https://github.com/taksau/GPS-Net.

In recent years, microblog systems such as Twitter and Sina Weibo have averaged multimillion active users. On the other hand, the microblog system has become a new means of rumor-spreading platform. In this paper, we investigate the machine-learning-based rumor identification approaches. We observed that feature design and selection has a stronger impact on the rumor identification accuracy than the selection of machine-learning algorithms. Meanwhile, the rumor publishers' behavior may diverge from normal users', and a rumor post may have different responses from a normal post. However, mass behavior on rumor posts has not been explored adequately. Hence, we investigate rumor identification schemes by applying five new features based on users' behaviors, and combine the new features with the existing well-proved effective user behavior-based features, such as followers' comments and reposting, to predict whether a microblog post is a rumor. Experiment results on real-world data from Sina Weibo demonstrate the efficacy and efficiency of our proposed method and features. From the experiments, we conclude that the rumor detection based on mass behaviors is more effective than the detection based on microblogs' inherent features.