The complexity and diversity of buildings make it challenging to extract low-level and high-level features with strong feature representation by using deep neural networks in building extraction tasks. Meanwhile, deep neural network-based methods have many network parameters, which take up a lot of memory and time in training and testing. We propose a novel fully convolutional neural network called the Context Feature Enhancement Network (CFENet) to address these issues. CFENet comprises three modules: the spatial fusion module, the focus enhancement module, and the feature decoder module. First, the spatial fusion module aggregates the spatial information of low-level features to obtain buildings’ outline and edge information. Secondly, the focus enhancement module fully aggregates the semantic information of high-level features to filter the information of building-related attribute categories. Finally, the feature decoder module decodes the output of the above two modules to segment the buildings more accurately. In a series of experiments on the WHU Building Dataset and the Massachusetts Building Dataset, our CFENet balances efficiency and accuracy compared to the other four methods we compared, and achieves optimality on all five evaluation metrics: PA, PC, F1, IoU, and FWIoU. This indicates that CFENet can effectively enhance and fuse buildings’ low-level and high-level features, improving building extraction accuracy.

The rise of deep learning technology has broadly promoted the practical application of artificial intelligence in production and daily life. In computer vision, many human-centered applications, such as video surveillance, human-computer interaction, digital entertainment, etc., rely heavily on accurate and efficient human pose estimation techniques. Inspired by the remarkable achievements in learning-based 2D human pose estimation, numerous research studies are devoted to the topic of 3D human pose estimation via deep learning methods. Against this backdrop, this paper provides an extensive literature survey of recent literature about deep learning methods for 3D human pose estimation to display the development process of these research studies, track the latest research trends, and analyze the characteristics of devised types of methods. The literature is reviewed, along with the general pipeline of 3D human pose estimation, which consists of human body modeling, learning-based pose estimation, and regularization for refinement. Different from existing reviews of the same topic, this paper focus on deep learning-based methods. The learning-based pose estimation is discussed from two categories: single-person and multi-person. Each one is further categorized by data type to the image-based methods and the video-based methods. Moreover, due to the significance of data for learning-based methods, this paper surveys the 3D human pose estimation methods according to the taxonomy of supervision form. At last, this paper also enlists the current and widely used datasets and compares performances of reviewed methods. Based on this literature survey, it can be concluded that each branch of 3D human pose estimation starts with fully-supervised methods, and there is still much room for multi-person pose estimation based on other supervision methods from both image and video. Besides the significant development of 3D human pose estimation via deep learning, the inherent ambiguity and occlusion problems remain challenging issues that need to be better addressed.

Geometrical structures and the internal local region relationship, such as symmetry, regular array, junction, etc., are essential for understanding a 3D shape. This paper proposes a point cloud feature extraction network named PointSCNet, to capture the geometrical structure information and local region correlation information of a point cloud. The PointSCNet consists of three main modules: the space-filling curve-guided sampling module, the information fusion module, and the channel-spatial attention module. The space-filling curve-guided sampling module uses Z-order curve coding to sample points that contain geometrical correlation. The information fusion module uses a correlation tensor and a set of skip connections to fuse the structure and correlation information. The channel-spatial attention module enhances the representation of key points and crucial feature channels to refine the network. The proposed PointSCNet is evaluated on shape classification and part segmentation tasks. The experimental results demonstrate that the PointSCNet outperforms or is on par with state-of-the-art methods by learning the structure and correlation of point clouds effectively.

Feature selection is a multi-objective problem, which can eliminate irrelevant and redundant features and improve the accuracy of classification at the same time. Feature selection is a great challenge to balance the conflict between the two goals of selection accuracy and feature selection ratio. The evolutionary algorithm has been proved to be suitable for feature selection. Recently, a new meta-heuristic algorithm named the crow search algorithm has been applied to the problem of feature selection. This algorithm has the advantages of few parameters and achieved good results. However, due to the lack of diversity in late iterations, the algorithm falls into local optimal problems. To solve this problem, we propose the adaptive hierarchical learning crow search algorithm (AHL-CSA). Firstly, an adaptive hierarchical learning technique was used to adaptive divide the crow population into several layers, with each layer learning from the top layer particles and the topmost layer particles learning from each other. This strategy encourages more exploration by lower individuals and more exploitation by higher individuals, thus improving the diversity of the population. In addition, in order to make full use of the search information of each level in the population and reduce the impact of local optimization on the overall search performance of the algorithm, we introduce an information sharing mechanism to help adjust the search direction of the population and improve the convergence accuracy of the algorithm. Finally, different difference operators are used to update the positions of particles at different levels. The diversity of the population is further improved by using different difference operators. The performance of the method was tested on 18 standard UCI datasets and compared with eight other representative algorithms. The comparison of experimental results shows that the proposed algorithm is superior to other competitive algorithms. Furthermore, the Wilcoxon rank-sum test was used to verify the validity of the results.

Feature selection is a typical multiobjective problem including two conflicting objectives. In classification, feature selection aims to improve or maintain classification accuracy while reducing the number of selected features. In practical applications, feature selection is one of the most important tasks in remote sensing image classification. In recent years, many metaheuristic algorithms have attempted to explore feature selection, such as the dragonfly algorithm (DA). Dragonfly algorithms have a powerful search capability that achieves good results, but there are still some shortcomings, specifically that the algorithm’s ability to explore will be weakened in the late phase, the diversity of the populations is not sufficient, and the convergence speed is slow. To overcome these shortcomings, we propose an improved dragonfly algorithm combined with a directed differential operator, called BDA-DDO. First, to enhance the exploration capability of DA in the later stages, we present an adaptive step-updating mechanism where the dragonfly step size decreases with iteration. Second, to speed up the convergence of the DA algorithm, we designed a new differential operator. We constructed a directed differential operator that can provide a promising direction for the search, then sped up the convergence. Third, we also designed an adaptive paradigm to update the directed differential operator to improve the diversity of the populations. The proposed method was tested on 14 mainstream public UCI datasets. The experimental results were compared with seven representative feature selection methods, including the DA variant algorithms, and the results show that the proposed algorithm outperformed the other representative and state-of-the-art DA variant algorithms in terms of both convergence speed and solution quality.

Self-attention networks have revolutionized the field of natural language processing and have also made impressive progress in image analysis tasks. Corrnet3D proposes the idea of first obtaining the point cloud correspondence in point cloud registration. Inspired by these successes, we propose an unsupervised network for non-rigid point cloud registration, namely NrtNet, which is the first network using a transformer for unsupervised large deformation non-rigid point cloud registration. Specifically, NrtNet consists of a feature extraction module, a correspondence matrix generation module, and a reconstruction module. Feeding a pair of point clouds, our model first learns the point-by-point features and feeds them to the transformer-based correspondence matrix generation module, which utilizes the transformer to learn the correspondence probability between pairs of point sets, and then the correspondence probability matrix conducts normalization to obtain the correct point set corresponding matrix. We then permute the point clouds and learn the relative drift of the point pairs to reconstruct the point clouds for registration. Extensive experiments on synthetic and real datasets of non-rigid 3D shapes show that NrtNet outperforms state-of-the-art methods, including methods that use grids as input and methods that directly compute point drift.