Abstract Landslides are a type of sudden and highly destructive geological hazard. Traditional detection methods often suffer from delayed response and low efficiency. In recent years, deep learning-based object detection techniques have attracted increasing attention in disaster recognition tasks, particularly transformer-based detection models, which exhibit significant advantages in global feature modeling. However, landslide targets in remote sensing imagery often present challenges such as large-scale variation, blurred boundaries, and texture interference. To address these issues, this study proposes an improved detection algorithm based on the RT-DETR-r18 framework by integrating multiple specialized modules. First, the DDC3 module is designed to enhance the recognition of fine boundaries and local textures, thereby improving the feature extraction capacity of the backbone network. Second, an Efficient Additive Attention (EAA) mechanism is introduced to suppress redundant information and strengthen the model's focus on critical regions, improving detection precision. Finally, the CGAFusion module is employed, which utilizes a triple-attention strategy to collaboratively regulate feature weights. This module enhances the model’s ability to filter salient features while preserving global contextual information, leading to more accurate landslide edge detection. A dual-class dataset comprising landslides and storms is constructed from multi-source imagery for evaluation. The experimental results show that the proposed method outperforms existing models in several dimensions including [email protected] and F1 score, demonstrating strong detection accuracy. Code is available at https://github.com/sanyauChenCoder/Landslide_02.git .

The assembly of complete and circularized mitochondrial genomes (mitogenomes) is essential for population genetics, phylogenetics and evolution studies. Recently, Song et al. developed a seed-free tool called MEANGS for de novo mitochondrial assembly from whole genome sequencing (WGS) data in animals, achieving highly accurate and intact assemblies. However, the suitability of this tool for marine fish remains unexplored. Additionally, we have concerns regarding the overlap sequences in their original results, which may impact downstream analyses. In this Letter to the Editor, the effectiveness of MEANGS in assembling mitogenomes of cartilaginous and ray-finned fish species was assessed. Moreover, we also discussed the appropriate utilization of MEANGS in mitogenome assembly, including the implementation of the data-cut function and circular detection module. Our observations indicated that with the utilization of these modules, MEANGS efficiently assembled complete and circularized mitogenomes, even when handling large WGS datasets. Therefore, we strongly recommend users employ the data-cut function and circular detection module when using MEANGS, as the former significantly reduces runtime and the latter aids in the removal of overlapped sequences for improved circularization. Furthermore, our findings suggested that approximately 2× coverage of clean WGS data was sufficient for MEANGS to assemble mitogenomes in marine fish species. Moreover, due to its seed-free nature, MEANGS can be deemed one of the most efficient software tools for assembling mitogenomes from animal WGS data, particularly in studies with limited species or genetic background information.

In this paper, we aim to reconstruct a full 3D human shape from a single image. Previous vertex-level and parameter regression approaches reconstruct 3D human shape based on a pre-defined adjacency matrix to encode positive relations between nodes. The deep topological relations for the surface of the 3D human body are not carefully exploited. Moreover, the performance of most existing approaches often suffer from domain gap when handling more occlusion cases in real-world scenes. In this work, we propose a Deep Mesh Relation Capturing Graph Convolution Network, DC-GNet, with a shape completion task for 3D human shape reconstruction. Firstly, we propose to capture deep relations within mesh vertices, where an adaptive matrix encoding both positive and negative relations is introduced. Secondly, we propose a shape completion task to learn prior about various kinds of occlusion cases. Our approach encodes mesh structure from more subtle relations between nodes in a more distant region. Furthermore, our shape completion module alleviates the performance degradation issue in the outdoor scene. Extensive experiments on several benchmarks show that our approach outperforms the previous 3D human pose and shape estimation approaches.

We propose a method to extract representative features for fashion analysis by utilizing weakly annotated online fashion images in this work. The proposed system consists of two stages. In the first stage, we attempt to detect clothing items in a fashion image: the top clothes (t), bottom clothes (b) and one-pieces (o). In the second stage, we extract discriminative features from detected regions for various applications of interest. Unlike previous work that heavily relies on well-annotated fashion data, we propose a way to collect fashion images from online resources and conduct automatic annotation on them. Based on this methodology, we create a new fashion dataset, called the Web Attributes, to train our feature extractor. It is shown by experiments that extracted regional features can capture local characteristics of fashion images well and offer better performance than previous works.

One fundamental problem of the convolutional neural network (CNN) is catastrophic forgetting, which occurs when new object classes and data are added while the original dataset is not available any more. Training the network only using the new dataset deteriorates the performance with respect to the old dataset. To overcome this problem, we propose an expanded network architecture, called the ExpandNet, to enhance the CNN incremental learning capability. Our solution keeps filters of the original networks on one hand, yet adds additional filters to the convolutional layers as well as the fully connected layers on the other hand. The proposed new architecture does not need any information of the original dataset, and it is trained using the new dataset only. Extensive evaluations based on the CIFAR -10 and the CIFAR -100 datasets show that the proposed method has a slower forgetting rate as compared to several existing incremental learning networks.

An efficient, scalable and robust approach to the handwritten digits recognition problem based on the Saak transform is proposed in this work. First, multi-stage Saak transforms are used to extract a family of joint spatial-spectral representations of input images. Then, the Saak coefficients are used as features and fed into the SVM classifier for the classification task. In order to control the size of Saak coefficients, we adopt a lossy Saak transform that uses the principal component analysis (PCA) to select a smaller set of transform kernels. The handwritten digits recognition problem is well solved by the convolutional neural network (CNN) such as the LeNet-5. We conduct a comparative study on the performance of the LeNet-5 and the Saak-transform-based solutions in terms of scalability and robustness as well as the efficiency of lossless and lossy Saak transforms under a comparable accuracy level.