Cross-and-Diagonal Networks: An Indirect Self-Attention Mechanism for Image Classification.

In recent years, computer vision has witnessed remarkable advancements in image classification, specifically in the domains of fully convolutional neural networks (FCNs) and self-attention mechanisms. Nevertheless, both approaches exhibit certain limitations. FCNs tend to prioritize local information, potentially overlooking crucial global contexts, whereas self-attention mechanisms are computationally intensive despite their adaptability. In order to surmount these challenges, this paper proposes cross-and-diagonal networks (CDNet), innovative network architecture that adeptly captures global information in images while preserving local details in a more computationally efficient manner. CDNet achieves this by establishing long-range relationships between pixels within an image, enabling the indirect acquisition of contextual information. This inventive indirect self-attention mechanism significantly enhances the network's capacity. In CDNet, a new attention mechanism named "cross and diagonal attention" is proposed. This mechanism adopts an indirect approach by integrating two distinct components, cross attention and diagonal attention. By computing attention in different directions, specifically vertical and diagonal, CDNet effectively establishes remote dependencies among pixels, resulting in improved performance in image classification tasks. Experimental results highlight several advantages of CDNet. Firstly, it introduces an indirect self-attention mechanism that can be effortlessly integrated as a module into any convolutional neural network (CNN). Additionally, the computational cost of the self-attention mechanism has been effectively reduced, resulting in improved overall computational efficiency. Lastly, CDNet attains state-of-the-art performance on three benchmark datasets for similar types of image classification networks. In essence, CDNet addresses the constraints of conventional approaches and provides an efficient and effective solution for capturing global context in image classification tasks.

Fast reconstruction technology of a laser beam spatial transmission characteristic curve.

The traditional multiposition method of the M2 factor measurement system is a good method, but it is relatively time consuming, so it cannot meet the requirements of the transient test of a Gaussian beam. To solve this problem, a quadriwave lateral shearing interferometer and a wavefront construction method based on a difference Zernike polynomial are analyzed. This interferometer uses a special grating to select four replicas of the wavefront, and the interferogram generated by four replicas includes difference wavefront information. Then the difference Zernike polynomial method is used to analyze quadriwave lateral shearing interferograms. The characteristic parameters are obtained after finding the optimal terms of the Zernike polynomials. As a result, the errors of F-number, beam radius, and radius of curvature are 3.7%, 3.8%, and 0.6%, respectively, which verifies this method to calculate parameters of Gaussian beam. In addition, we also find that the shear amount has influence on the reconstructed wavefront. It is showed that as the amount of shear increases from 20 pixels to 90 pixels, the peak-to-valley (P-V) values and RMS values both gradually decrease with a nonlinear relationship, which could be used to decrease the error of wavefront reconstruction further.

PointCartesian-Net: enhancing 3D coordinates for semantic segmentation of large-scale point clouds.

Collecting accurate outdoor point cloud data depends on complex algorithms and expensive experimental equipment. The requirement of data collecting and the characteristics of point clouds limit the development of semantic segmentation technology in point clouds. Therefore, this paper proposes a neural network model named PointCartesian-Net that uses only 3D coordinates of point cloud data for semantic segmentation. First, to increase the feature information and reduce the loss of geometric information, the 3D coordinates are encoded to establish a connection between neighboring points. Second, a dense connect and residual connect are employed to progressively increase the receptive field for each 3D point, and aggregated multi-level and multi-scale semantic features obtain rich contextual information. Third, inspired by the success of the SENet model in 2D images, a 3D SENet that learns the relation between the characteristic channels is proposed. It allows the PointCartesian-Net to weight the informative features while suppressing less useful ones. The experimental results produce 60.2% Mean Intersection-over-Union and 89.1% overall accuracy on the large-scale benchmark Semantic3D dataset, which shows the feasibility and applicability of the network.

New discrete heavy tailed distributions as models for insurance data.

Although many data sets are discrete and heavy tailed (for example, number of claims and claim amounts if recorded as rounded values), not many discrete heavy tailed distributions are available in the literature. In this paper, we discuss thirteen known discrete heavy tailed distributions, propose nine new discrete heavy tailed distributions and give expressions for their probability mass functions, cumulative distribution functions, hazard rate functions, reversed hazard rate functions, means, variances, moment generating functions, entropies and quantile functions. Tail behaviour and a measure of asymmetry are used to compare the known and new discrete heavy tailed distributions. The better fits of the discrete heavy tailed distributions over their continuous counterparts as assessed by probability plots are illustrated using three data sets. Finally, a simulated study is performed to assess the finite sample performance of the maximum likelihood estimators used in the data application section.