Jianzhi Yu

Glacier change detection is vital for understanding climate dynamics and assessing environmental risks. However, existing change detection methods that indiscriminately model local details (change edges) and global context (change bodies) often yield incomplete edges and inconsistent interiors. To address this issue, we propose a dual-stream change feature disentanglement network (DCFDNet). At its core, the change feature disentanglement module (CFDM) explicitly disentangles change features into body and edge components using a learnable flow field. Building on this, the dual-stream change feature refinement (DCFR) scheme enhances the disentangled features through two complementary streams: the body consistency stream leverages large-kernel attention to enforce global spatial consistency of change bodies, while the edge integrity stream applies local graph modeling to sharpen fine-grained change edges. The fusion of both streams produces a unified change representation that maintains both body consistency and edge integrity. Experiments on a polar glacier dataset demonstrate that DCFDNet outperforms recent state-of-the-art methods, delivering coherent body regions and accurate edge localization under challenging polar conditions.

Monitoring glacier changes is essential for understanding global climate dynamics and assessing their environmental impacts. However, accurate detection remains challenging due to seasonal variations, illumination differences, and heterogeneous textures in remote sensing imagery. To address these issues, we propose VPGCD-Net, a Transformer-based dual-branch network that achieves robust glacier change detection through visual prompt engineering. The visual prompting branch integrates threshold segmentation and difference calculation, leveraging a visual prompt transformer (VPT) to encode regions of significant change and generate high-level semantic prompts. Meanwhile, the change detection branch adopts ResNet18 as the backbone to extract dual-temporal features, followed by a Transformer module for modeling global spatiotemporal dependencies and a FiLM module for adaptive feature modulation to emphasize real change regions. Complementing the method, we introduce the first polar glacier-focused dataset specifically designed for deep learning-based glacier change detection in remote sensing. Experimental results demonstrate that VPGCD-Net outperforms existing state-of-the-art methods, achieving superior accuracy even under complex conditions such as shadow interference. The dataset is publicly available at https://huggingface.co/datasets/cuibinge/Glacier-Dataset.

Domain adaptation methods enable model migration and adaptation across different domain data distributions. However, the source and target domains of hyperspectral images (HSIs) have large spectral offsets and spatial distribution differences, making the extraction of high-quality domain-invariant features between different domains is essential for classification. To achieve a more consistent feature representation for each category between the source and target domains, we propose a category semantic guided unsupervised domain adaptation network (CSGNet) for HSIs classification. CSGNet is designed to learn cross-domain invariant representation from category semantic information. First, to embed category semantic prior knowledge during feature learning, we extracted textual semantic features from the textual descriptions for each category and projected visual features into the semantic space via visual-linguistic alignment. A category representation memory pool is then introduced to store the visual-linguistic representations of different categories. Second, we propose a bi-classifier adversarial learning method designed to generate inconsistent category predictions in the unlabeled target domain, thereby enhancing the classifier’s discriminative capability regarding those hard-to-transfer features. Finally, to utilize the domain-invariant features stored in the category memory pool, a category attention module is proposed to guide the model’s adaptation to the data from different domains, mitigating the impact of the differences in the domain data distributions. Extensive experimental results validated on three cross-domain datasets demonstrate that the proposed method outperforms other state-of-the-art methods. The source code is available at http://github.com/cuibinge/CSGNet.

The precise identification of semantic changes in remote sensing images is of great significance in the domains of urban planning and disaster assessment. Nevertheless, current change detection models are inadequate when it comes to model semantic interactions in pairings of temporal images. This leads to poor identification of identical semantic targets that have unique features. In this paper, we put forth a proposal for a Multilevel Intertemporal Semantic Guidance Network (MISGNet) that would effectively derive representations of semantic changes. The bitemporal images' multilevel features are initially extracted using a transformer feature extractor. These extracted features are then bi-directionally semantically augmented using a Semantic Guidance Module (SGM) in order to acquire more comprehensive semantic representations. Particularly, in order to obtain the object's semantic representation, the land cover objects in the multilevel features of the bitemporal images are soft clustered and mapped to the graph space, with each vertex representing an object in the graph. Then, bidirectional semantic enhancement is achieved through the use of intertemporal non-local operations, which strengthen the semantic representation of the bitemporal images. Moreover, a Multilevel Difference Aggregation Module (MDAM) is implemented in order to enhance the efficacy of summarizing distinctions between different levels and highlighting semantic changes in features. This is achieved through the utilization of pixel-wise addition and pixel-wise multiplication. Extensive experiments conducted on three publicly available datasets provide irrefutable evidence that our model outperforms alternative methods across various evaluation metrics.

Radiology report generation that aims to accurately describe medical findings for given images, is pivotal in contemporary computer-aided diagnosis. Recently, despite considerable progress, current radiology report generation models still struggled to achieve consistent quality across difficult and easy samples, which dramatically impacts their clinical value. To solve this problem, we explore the difficult samples mining in radiology report generation and propose the Linear Hybrid-Reward based Reinforced Focal Learning (LHR-RFL) to effectively guide the model to allocate more attention towards some difficult samples, thereby enhancing its overall performance in both general and intricate scenarios. In implementation, we first propose the Linear Hybrid-Reward (LHR) module to better quantify the learning difficulty, which employs a linear weighting scheme that assigns varying weights to three representative Natural Language Generation (NLG) evaluation metrics. Then, we propose the Reinforced Focal Learning (RFL) to adaptively adjust the contributions of difficult samples during training, thereby augmenting their impact on model optimization. The experimental results demonstrate that our proposed LHR-RFL improves the performance of the base model across all NLG evaluation metrics, achieving an average performance improvement of 20.9% and 13.2% on IU X-ray and MIMIC-CXR datasets, respectively. Further analysis also proves that our LHR-RFL can dramatically improve the quality of reports for difficult samples. The source code will be available at https://github.com/SKD-HPC/LHR-RFL.

Significant progress has been made in modern remote sensing (RS) image change detection (CD) by leveraging the powerful feature learning capabilities of convolutional neural networks (CNNs) and transformers. However, current popular change detection techniques primarily focus on extracting deep semantic features and pixel-level interactions while overlooking the potential benefits of cluster-level semantic interaction in bitemporal images. In this letter, we propose a novel approach called the Bitemporal Graph Semantic Interaction Network for Remote Sensing Images Change Detection (BGSINet-CD). Specifically, the land cover types in bitemporal images are clustered by employing soft clustering for each pixel, and then each cluster is separately projected to a vertex in graph space. Additionally, we introduce a graph semantic interaction module (GSIM) that enhances the interactions between bitemporal features at the semantic level. GSIM effectively improves the information coupling between bitemporal features, thereby suppressing task-irrelevant information. In comparison to other competing methods, our approach demonstrates a significant improvement in F1 scores, achieving 88.25% and 91.02% on the GZ-CD and WHU-CD datasets, respectively. Furthermore, our method employs a reduced number of network parameters and exhibits lower complexity, striking a superior balance between accuracy and computational efficiency.