In the era of increasingly scarce spectrum resources, electromagnetic spectrum (EMS) prediction has emerged as a critical means for enhancing spectrum utilization efficiency. However, most of the existing EMS methods primarily exploit low-dimensional features such as temporal, frequency, or spatial characteristics in an individual fashion, which limits their ability to fully capture the inherent complexity of spectrum dynamics. To improve the performance, this paper proposes a novel EMS prediction model, which involving three operations, namely multi-dimensional decoupling, feature fusion and temporal prediction. Firstly, for multi-dimensional decoupling operation, we propose a Multi-dimensi... More >

Multimodal Sentiment Analysis (MSA), a pivotal task in affective computing, aims to enhance sentiment understanding by integrating heterogeneous data from modalities such as text, images, and audio. However, existing methods continue to face challenges in semantic alignment, modality fusion, and interpretability. To address these limitations, we propose VBCSNet, a hybrid attention-based multimodal framework that leverages the complementary strengths of Vision Transformer (ViT), BERT, and CLIP architectures. VBCSNet employs a Structured Self-Attention (SSA) mechanism to optimize intra-modal feature representation and a Cross-Attention module to achieve fine-grained semantic alignment across m... More >

The need for robust information protection techniques has become important in security applications, especially to safeguard secret messages during transmission. The rapid expansion and pervasive use of the Internet have amplified security concerns, particularly regarding the authenticity of digital images, a significant issue in the context of the fourth industrial revolution. Biometric authentication using image watermarking addresses these concerns by embedding biometric information into digital images, thereby ensuring their security and privacy. Numerous recent methods have fused biometric modalities with watermarking techniques to enhance the security and reliability of transmitted mes... More >

Real-time detection of small obstacles is a critical challenge for autonomous systems such as self-driving vehicles, unmanned aerial vehicles (UAVs), and mobile robots. These small obstacles (e.g., road debris, fallen branches, cables) pose significant safety risks due to their low visibility and irregular appearances. This paper presents a comprehensive systematic review of 117 technical articles published between 2016 and 2025, focusing on the techniques and deployment strategies for real-time small obstacle detection using fused multidimensional information. We summarize and analyze developments in small obstacle definitions, sensing hardware, detection algorithms, fusion methods, and rea... More >

This research addresses the challenge of detecting radar range-spread targets in compound-Gaussian clutter environments. In such scenarios, the target signals occupy unknown coordinates within a subspace, while the clutter is modeled as compound-Gaussian distribution involving inverse Gamma textures and complex Gaussian speckles with an unknown persymmetric covariance matrix. Additionally, we assume the availability of training data for estimating clutter covariance matrix. Utilizing a two-step approach, three detectors are proposed according to the Gradient, Rao, and Wald tests, by taking advantage of the persymmetry of the clutter covariance matrix. Theoretical analyses demonstrate that th... More >

With the continuous expansion of the application field of artificial intelligence, radar data processing has also begun to fully enter the era of intelligence, and new achievements have emerged in the research fields of target detection, target tracking, and target recognition. At a time of rapid development of artificial intelligence technology, it is necessary to think about the future development of radar data intelligent processing. To this end, combining the research history and the superficial understanding of radar data intelligent processing in the past ten years, our team analyzes the main research progress and challenges of radar data intelligent processing, examines the new requir... More >

With the rise of data engineering-driven automatic annotation strategies, deep learning has demonstrated remarkable performance and strong competitiveness in intelligent fault diagnosis. However, the inherent limitations of automatic annotators inevitably introduce noisy labels, which in turn hinder the generalization and accuracy of diagnostic models. Although numerous Learning with Noisy Labels (LNL) methods attempt to alleviate the impact of label noise through sample selection or label correction, most rely heavily on model predictions to guide training. This self-reinforcing mechanism frequently leads to confirmation bias, especially under high-noise conditions, thereby limiting their e... More >

Considering the large-area distribution, smooth brightness gradients, and blurred boundaries of Mura defects in real industrial scenarios, as well as the challenge of balancing accuracy and efficiency in existing methods, we propose a lightweight deep learning-based detection method for large-area Mura defects, termed SIFNet. The SIFNet adopts a classical encoder-decoder architecture with MobileNet-V2 as the backbone. Furthermore, we design a Graph-based Semantic Interscale-fusion Block (GSIB) that integrates the Semantic Fluid Aggregation Module (SFAM) and the Semantic Graph Inference Module (SGIM) to collaboratively extract high-level semantic features across multiple scales and establish... More >