Wood surface defect inspection is a key step in production-line quality control. However, supervised detectors typically require large-scale annotations and often generalize poorly to unseen defect types. In this paper, we formulate wood defect inspection as an unsupervised anomaly detection problem and propose CoMemRet, a framework based on multiview contrastive learning and memory bank retrieval. CoMemRet trains a multiview contrastive encoder in the Lab color space to decouple luminance structure from chromatic variation, and introduces a patch-level contrastive constraint to enhance representations of fine-grained texture anomalies. The extracted features are further refined by feature gating and fusion modules to obtain more stable and compact embeddings. During anomaly detection, CoMemRet constructs a memory bank from a coreset of normal features and computes anomaly scores from the k-NN distances between test features and the normal-feature coreset. A clustering-guided coreset subsampling strategy is further introduced to improve the efficiency of coreset construction and k-NN retrieval while preserving coverage of normal patterns. Experiments on the newly collected Wood Surface Anomaly Detection Dataset (WSAD) and the public MVTec AD (wood) benchmark demonstrate competitive performance and improved generalization, suggesting suitability for production-line inspection.

Interactive medical image segmentation plays a crucial role in medical diagnosis, treatment planning, and in-terventional procedures. With the rapid advancement of artificial intelligence and deep learning technologies, this field has not only seen significant improvements in the accuracy and efficiency of segmenting anatomical structures and pathological areas from medical images, but it has also made substantial strides in the development of user-centric interactive tools. The introduction of the Segmentation Anything Model (SAM) represents a significant expansion of prompt-driven approaches within the domain of image segmentation, introducing a plethora of previously untapped functionalities. However, due to the substantial differences between natural and medical images, the generalization capabilities of interactive segmentation models across different imaging techniques remain limited, necessitating the ongoing development of targeted interactive segmentation models. In this work, we provide a comprehensive overview aimed at extending the efficacy of SAM to various medical imaging modalities, encompassing diverse image analyses and domains. Additionally, we explore potential research directions for SAM in different medical imaging fields. Despite these advancements, the accuracy and robustness of segmentation models still largely depend on high-quality annotated data, which is costly and time-consuming to acquire. These research efforts herald significant forthcoming advancements in interactive medical image segmen-tation technology, with the potential to greatly improve patient outcomes and streamline medical procedures, thereby driving further development and innovation in the healthcare industry.

Accurate and reproducible assessment of myocardial conditions is essential for the diagnosis of prior infarctions, cardiomyopathies, and inflammatory diseases. Although cardiac magnetic resonance imaging (CMR) is regarded as the gold standard for evaluating myocardial anatomy and function, manual segmentation remains labor-intensive and susceptible to variability. The demand for high-resolution images, dense predictions, and comprehensive segmentation across all phases of the cardiac cycle introduces significant challenges for Cine-CMR analysis. To address these issues, we present EML-SAM, an innovative interactive segmentation model based on the Segment Anything Model (SAM). EML-SAM incorporates an early-mid-late (EML) multi-scale fusion strategy, effectively mitigating the premature dilution of interaction information, thereby reducing segmentation errors and enhancing feature utilization. Furthermore, the model utilizes a multi-scale linear attention (MLA) transformer to establish an efficient global receptive field. Comprising 12 Transformer layers, the model integrates multi-scale attention with multi-scale perceptrons, striking a balance between computational efficiency and capacity. The proposed methodology offers a robust and efficient solution for high-resolution Cine-CMR segmentation, delivering accurate and real-time segmentation performance throughout the various stages of temporal changes within the cardiac cycle.

Fabric defect detection is critical for quality control in the textile industry. However, current challenges include the scarcity of high-quality datasets, the difficulty of detecting small defects in complex environments, and suboptimal detection performance for elongated defects. To address these issues, we propose an improved Real-Time Detection Transformer (RT-DETR) model specifically designed for fabric defect detection in industrial settings. A custom fiber defect dataset was created with random augmentations for improved diversity. We designed the Global Attention-based Instance Feature Interaction (GA-AIFI) module to enhance local feature extraction, improving the model's detection capability for small defects. Additionally, to address the prevalence of elongated defects in industrial fabric production, we introduced the Aspect Ratio-aware Distance-IoU (AR-DIoU) loss function, which further improves bounding box localization accuracy. Experiments conducted on both the custom dataset and the MVTec dataset demonstrate significant improvements, with [email protected] increasing by 3.5% and 3.6%,respectively, confirming the model's effectiveness.

Fabric defect detection is a critical research area in the textile industry, with substantial practical implications. However, existing methods often struggle with detecting small and multi-scale defects. To address these challenges, we propose NR-DETR, a lightweight real-time detection model built on an enhanced RT-DETR framework. The model employs multi-scale enhancement and logical distillation to boost detection performance. Specifically, the RB-AIFI and CSCB modules are designed to optimize the stability of multi-scale feature fusion and enhance detection capabilities, while the efficient upsampling module (EUCB) significantly boosts inference efficiency. Furthermore, the model incorporates logical and feature distillation methods, employing hierarchical feature alignment and a shared decoder design to enhance the expressiveness and detection accuracy of the student model. Experimental results on the MVTec Fabric Defect dataset demonstrate that NR-DETR achieves a detection accuracy of 79.2%, a 5.2% improvement over the baseline model, while reducing the parameter count by approximately 10%, showcasing superior performance and efficiency. After applying the proposed joint distillation strategy, detection accuracy further improves by 1.7%, validating the effectiveness of the approach. The proposed model offers robust technical support for real-time and efficient fabric defect detection tasks.

In the fault diagnosis of industrial equipment, transfer learning alleviates the problem of data distribution offset and annotation scarcity through cross domain knowledge migration. However, the existing methods have limitations. Single scale feature alignment ignores the difference between shallow and deep features. The pseudo label strategy with fixed temperature parameters reduces the feature discrimination, and the single distribution alignment is difficult to take into account the global and local structure. Therefore, this paper proposes a multi-scale dynamic confrontation transfer learning framework (MDATL), which includes a hierarchical dynamic confrontation mechanism, and dynamically adjusts the characteristic confrontation intensity of each layer through periodic GRL; the two-stage pseudo label optimization strategy combined with temperature scaling softens the probability distribution of the target domain; a mixed distribution alignment strategy, which combines MMD and CORAL to dynamically balance global statistics and local covariance. Experiments using CWRU and PU data sets of six groups of cross condition task verification show the superior performance of this method in unsupervised fault diagnosis.

Addressing the challenges posed by the insufficient computational power of low-spec devices to achieve the inference efficiency of prevailing deep learning models, alongside the variability in type and shape within industrial metal sheet datasets, while also catering to the demand for high-precision detection, this paper proposes a novel single-stage detection architecture termed MTRNet.In this paper, we introduce three key enhancements: Firstly, to reduce the number of parameters, we propose a novel Partial Depth Convolution (PDC) structure.By minimizing redundant computations, we aim to enhance the efficiency of spatial feature extraction.Secondly, we propose a Convolution Transformer (CTR) structure that replaces the self-attention module with a convolutional module.This modification addresses the computational inefficiency inherent in the self-attention mechanism of existing Visual Transformers (ViT).Finally, we introduce a novel Attention Convolution Transformer (ACTR) structure to enhance the extraction of global and local feature information.This architecture seamlessly integrates the strengths of both attention and convolution mechanisms, working synergistically to improve performance.Our proposed MTRNet network demonstrates superior detection performance compared to existing industrial standards, achieving a detection accuracy of 81.5% on the NEU-DET dataset.

Defects such as scratches, patches, and cracks frequently occur during sheet metal production. However, the low detection accuracy and slow processing speed of industrial defect detection models significantly impede enterprise production efficiency. The aforementioned issues primarily manifest in three aspects. Firstly, the model complexity and computational overhead are substantial. Secondly, detecting small local defects poses a significant challenge. Thirdly, extracting global features, such as elongated scratches, proves to be difficult. To address these challenges, this paper introduces a novel network architecture called SATRNet. Firstly, within the model backbone, the STR module is devised. Through incorporation of the sparse self-attention method and the CNNs parallel vision Transformer model in the shallow layers, this module significantly enhances the model's capability to extract global features. Secondly, the SCATR module is designed in this paper. By substituting self-attention with the designed SCA soft attention as the token mixer, the module aims to enhance detection accuracy while reducing the number of parameters, thereby fundamentally addressing the problem of model complexity. Finally, this paper presents the GCD bottleneck convolution module. This module combines shallow and deep features, enabling the fusion of more information beneficial for detection, thereby achieving improved efficacy in capturing minute targets. Experiments demonstrate that SATRNet surpasses existing advanced models in detection accuracy on public datasets.

Summary This article proposes a security strategy based on chaos encryption for the transmission of robot data on cloud platforms. Based on chaos encryption technology, the data information collected by the robot is encrypted before transmission. After receiving the request, the cloud platform server uses chaos decryption technology to analyze the ciphertext information and obtain the original relevant data information. The security strategy proposed in this article is simple and feasible, and can resist all kinds of security attacks.