Remote sensing (RS) images change detection (CD) is essential for the surveillance and prevention of geohazards. Nevertheless, the current deep learning (DL)-based CD methods still face challenges such as pseudo changes, missed detections, and edge noise due to the inadequate research of temporal differences and inconsistent viewing angles between the dual-temporal images. In order to improve the perception of spatiotemporal variations and effectively manage complex motion in spatiotemporal data, this paper proposes a spatiotemporal multi-scale fusion and optical flow warping network (SMOW-Net). Initially, the internal fusion property of 3D convolution enables the simultaneous extraction and fusion of feature information in dual-temporal images. The spatiotemporal multi-scale feature encoder (SMFE) module is proposed to mitigate the semantic gap between low-level and high-level features. This module is designed to aggregate complementary feature information between each level through temporal and spatial independent processing and flexible temporal transposed convolutional layers. Furthermore, the optical flow warper (OFW) module is intended to improve the spatiotemporal dynamic modeling capability in order to manage complex motion data effectively, where a two-channel spatial deformation field is autonomously learned by the network to guide feature alignment. The performance advantage of our network over eleven state-of-the-art methods (SOTA) on the GVLM-CD, LEVIR-CD, WHU-CD, S2Looking, and LEVIR-CD+ datasets is validated by experimental results. Finally, we also introduce SMOW-Net-LW, a lightweight variant with significantly reduced model complexity, suitable for resource-constrained settings, while still achieving excellent performance. The code for this work is available at https://github.com/ChundeLiao/SMOW-Net .

Federated learning (FL) is well-suited for multimodal tasks due to its ability to protect privacy and support local training. However, the complexity of real-world sensor environments causes modality heterogeneity across clients. Some modalities may be missing altogether, making it difficult to construct a generalized global model. Existing multimodal federated learning methods often address modality-missing scenarios under simplified assumptions of modality heterogeneity, typically focusing on unimodal clients and modality-complete multimodal clients. Moreover, to mitigate performance degradation caused by missing modalities, some approaches assume the availability of auxiliary information at the server, which may be impractical in real-world scenarios. Therefore, we propose a novel heterogeneous multimodal Federated Learning with Mask-Restoration and Self-Guidance (FL-MRSG). The Mask-Restoration employs a masking strategy to simulate missing data during feature extraction, enabling the network to learn semantic features of missing modality. Furthermore, we introduce an innovative self-guidance mechanism that leverages the restored data as guidance information, enabling the network to distinguish between complete and missing data representations. In addition, we propose a personalized decoupled aggregation strategy to facilitate the collaborative training of a global model across heterogeneous modality clients. We extend the multimodal test set to arbitrary modality combinations to evaluate the robustness of the global model. Extensive experiments on MOSI and SIMS datasets demonstrate the effectiveness of the proposed FL-MRSG for arbitrary missing modalities.

Imbalanced image classification faces critical challenges in balancing the quality and diversity of synthetic minority samples. This article proposes the improved estimation distribution algorithm-based latent feature distribution evolution (MEDA_LUDE) algorithm, an evolutionary algorithm-assisted deep distribution learning framework that optimizes latent feature distributions through a multivariate Gaussian mixture (GM) assumption and a novel four-phase training strategy. We introduce a large-margin GM (L-GM) loss to dynamically model covariances for feature learning and design a MEDA that evolves latent features via a similarity-guided fitness function, thus enhancing diversity while preserving synthesis quality. Extensive experiments demonstrate significant improvements: MEDA_LUDE achieves 95.9% accuracy on MNIST (imbalanced ratio-IR:100), surpassing state-of-the-art methods by 1.26% on CIFAR-10. For industrial fabric defect data sets, it elevates accuracy by 1.45% on DHU-FD and 0.92% on ALIYUN-FD, especially with precision and G-mean improvements of 2.5% and 1.17%, respectively, on DHU-FD. Visualizations confirm that MEDA_LUDE generates minority samples with superior quality-diversity tradeoffs. The framework's success in real-world fabric defect classification underscores its practical value in addressing imbalanced learning challenges.

Most fine-grained visual recognition methods endeavor to directly locate discriminative regions in intricate environments, but tend to overlook the object’s holistic structure, which may lead to misclassification due to overemphasizing incorrect areas. In this paper, we propose a coarse-to-fine paradigm, which prioritizes locating holistic structural regions of the target object, followed by a gradual search to locate discriminative areas. Specifically, we first design the “look into object” module to locate the areas encompassing the target’s holistic structure using prior information. Subsequently, without introducing additional parameters, we design a partial focus searching module to enhance feature representations of discriminative regions within the target’s structural composition. Ultimately, we segregate the foreground components from the original image, attaining a more precise characterization of the target. Furthermore, we demonstrate the practical application potential of our model in real-world industries through our self-constructed DHU-Fine-grained-6000 dataset. Comparative experiments on three public datasets indicate that the superiority of our approach over many recent methods and holds promising application potential in industrial production processes.

Multivariate time series prediction is a significant research area that aims to forecast future values based on past observations. Deep learning models with attention mechanisms have shown good predictive performance by emphasizing optimal-related sequences in the target series. However, these models ignore mutation information of nontarget sequences and the long short-term dependencies. To this end, a divide-and-rule combined learning method is proposed to address these limitations, which uses differentiated feature extractors to process different implicit features. First, we design a spatial and temporal information extractor to extract the time-dimensional feature information in the separation stage. Then, a multivariate mutation information extractor is constructed by convolution and maximum pooling layer to capture mutation information of nontarget sequences. Subsequently, the decoder component of the encoder-decoder model extracts long short-term dependencies while preserving the information of the target sequence to be predicted. Finally, in the cooperation stage, a feature fusion method based on a point attention mechanism is proposed, which can assign individual weights to each feature point and enhance the ability to focus on local areas. Experimental results on five real datasets in different domains show that the proposed method has better predictive performance compared to other baseline models.

In recent years, neural architecture search (NAS) methods have been proposed for the automatic generation of task-oriented network architecture in image classification. However, the architectures obtained by existing NAS approaches are optimized only for classification performance and do not adapt to devices with limited computational resources. To address this challenge, we propose a neural network architecture search algorithm aiming to simultaneously improve the network performance and reduce the network complexity. The proposed framework automatically builds the network architecture at two stages: block-level search and network-level search. At the stage of block-level search, a gradient-based relaxation method is proposed, using an enhanced gradient to design high-performance and low-complexity blocks. At the stage of network-level search, an evolutionary multiobjective algorithm is utilized to complete the automatic design from blocks to the target network. The experimental results demonstrate that our method outperforms all evaluated hand-crafted networks in image classification, with an error rate of 3.18% on Canadian Institute for Advanced Research (CIFAR10) and an error rate of 19.16% on CIFAR100, both at network parameter size less than 1 M. Obviously, compared with other NAS methods, our method offers a tremendous reduction in designed network architecture parameters.