Enhancing Few-Shot marble slab surface defect detection: A diffusion framework with knowledge distillation and semantic guidance

Deep learning-based surface defect segmentation approaches have shown promising performance in recent years. However, segmenting defects with complex shapes, large variations in size, and weakly textured defects with indistinct characteristics still poses significant challenges. In this article, a novel semantic guidance and texture priors based dual-branch surface defect segmentation network (SGTP-Net) is proposed for those issues. Firstly, we construct a feature extraction network combines semantic and texture branches. The semantic branch establishes global contextual relationships, while the texture branch captures local features of defects, this dual-branch ensured the network to extract features from various complex defects. Secondly, we design a feature fusion strategy based on semantic guidance and texture priors. The semantic information is used to guides the output of texture branch. After that, the guided texture information provides valuable edge texture priors for each layers output in semantic branch. The two branches mutually guide each other for improving ability of weak textures feature extraction. Finally, we run our method on the NEU-Seg, MT-Defect and MSD datasets to conduct a comprehensive comparison with some state-of-the-art general object segmentation models and specialized surface defect segmentation methods. The experimental results show that our SGTP-Net performs well in surface defect detection, offering excellent semantic segmentation accuracy and exhibiting good stability and robustness in detecting various surface defects.

Surface defect detection is a core component for ensuring the quality of multi-wedge belt products, yet traditional detection methods based on Convolutional Neural Networks (CNNs) suffer from weak global feature extraction and limited multi-scale feature fusion, severely impacting detection efficiency. Meanwhile, the constrained computational resources in industrial scenarios pose critical challenges to balancing model accuracy and speed. To address these issues, we propose a lightweight defect detection model, LCT-YOLO (Lightweight CNN-Transformer YOLO), which integrates the technical advantages of CNN and Transformer while meeting the lightweight deployment requirements of industrial applications. Specifically, a dynamic channel selection feature extraction network is constructed by introducing the partial channel selection module CSP_PTB to enhance the model’s capability of capturing long-distance semantic information. A lightweight multi-scale feature fusion network is developed by designing a multi-scale path aggregation network and integrating the dynamic upsampling module Dysample in the model’s neck, effectively addressing challenges from feature scale variations. A lightweight shared convolution detection head is introduced to tackle the high computational cost of existing model detection heads, optimizing performance and reducing model complexity through parameter sharing strategies. Finally, model pruning is employed to remove redundant parameters for compression, and knowledge distillation is utilized to compensate for accuracy loss. Experimental results show that the algorithm achieves an mAP50 of 94.3% on the dataset, with 0.8 M parameters and 2.0 GFLOPs computation, outperforming the baseline model by 0.9% in mAP50, reducing parameters by 74.2%, computation by 75.3%, and featuring a model size of only 1.9 M, making it highly suitable for edge device deployment. The algorithm surpasses other advanced methods in both accuracy and speed, not only meeting the dual requirements of detection precision and real-time performance in industrial scenarios but also being successfully applied to practical production, providing robust technical support for defect detection and product quality control.

Lightweight detection network for bridge defects based on model pruning and knowledge distillation

One of the greatest engineering feats in history is the construction of tunnels, and the management of tunnel safety depends heavily on the detection of tunnel defects. However, the real-time, portability, and accuracy issues with the present tunnel defect detection technique still exist. The study improves the traditional defect detection technology based on the knowledge distillation algorithm, the depth pooling residual structure is designed in the teacher network to enhance the ability to extract target features. Next, the MobileNetv3 lightweight network is built into the student network to reduce the number and volume of model parameters. The lightweight model is then trained in terms of both features and outputs using a multidimensional knowledge distillation approach. By processing the tunnel radar detection photos, the dataset is created. The experimental findings demonstrate that the multidimensional knowledge distillation approach greatly increases the detection efficiency: the number of parameters is decreased by 81.4%, from 16.03 MB to 2.98 MB, while the accuracy is improved by 2.5%, from 83.4% to 85.9%.

Due to the influence of construction quality, engineering geology and hydrological environment, defects such as dehollowing and insufficient compaction can occur in tunnels. Aiming at the problems of complex detection model, poor real-time performance and low accuracy of the current tunnel lining defect detection methods, the study proposes a lightweight defect detection algorithm of tunnel lining based on knowledge distillation. Firstly, a high-precision teacher model based on yolov5s was constructed by constructing a C3CSFM module that combines residual structure and attention mechanism, a MDFPN network structure with multi-scale feature fusion and a reweighted RWNMS re-screening mechanism. Secondly, in the distillation process, the feature and output dimension results are fused to improve the detection accuracy, and the mask feature relationship is learned in the space and channel dimension to improve the real-time detection. Tests on the tunnel lining radar defect image dataset showed that the number of parameters of the improved model was reduced from 16.03 MB to 3.20 MB, a reduction of 80%, and the average accuracy was improved from 83.4 to 86.5%, an increase of 3.1%. On the basis of maintaining the structure and detection performance of the model, the lightweight degree of the model is greatly improved, and the high-precision and real-time detection of tunnel lining defects is realized.

The classification method of steel surface defects with high performance and easy to be embedded in the detection equipment is one of the keys to ensure the quality of hot rolled strip. However, the development of deep convolutional neural networks (CNNs) in many real-world applications is largely hindered by their high computational cost, especially in industrial production, although it has good classification accuracy compared with machine learning-based methods in image recognition. Therefore, in this work, we present a lightweight network FCCNet based on the convolutional neural network to facilitate its application in the detection system. To compensate for the accuracy loss caused by the network downsizing, a knowledge distillation (KD) method using a larger trained network (teacher network) to teach a smaller network (student network) is adopted to improve the performance of our model. As a result, our method achieves a classification accuracy of 99.44%, precision of 99.46%, recall of 99.45%, and an F1 score of 99.45% on the NEU-CLS dataset, using only 0.03 MB parameters. These results show that the FCCNet is lighter than other existing classic CNNs with good performance for surface defects classification of hot-rolled steel strip, and it has the potential to be applied in the actual production line.

Fabric defect segmentation is an important part of ensuring the quality of product production. Using fabrics with surface defects will affect the quality and reputation of their products. In previous studies, some model compression methods have helped semantic segmentation models to be deployed on resource-limited working devices. However, the capacity reduction of models usually leads to a decline in detection performance. We propose a knowledge distillation method combining traditional KD loss and contrastive relational distillation, which makes student models learn the differential representation among various defects while receiving knowledge transfer from the teacher model. We use DeepLabV3+ and PSPNet with MobileNetV2 backbone networks as student models to validate our method. Experiment shows that our method performs better than direct training and traditional knowledge distillation methods on the DAGM and AITEX datasets. Our method enables lightweight models to achieve higher performance on fabric defect segmentation tasks.

In order to improve the defect recognition efficiency of transmission lines, the industry is currently using aerial images for automatic visual defect detection to ensure the safe operation of transmission lines. This paper proposes a method for defect recognition from coarse to fine, based on convolutional neural networks and connected domain algorithms, to improve recognition accuracy. The recognition speed is improved by using the knowledge distillation method of target detection networks based on decoupled features, adversarial features, and attention features. It has been found that the optimized recognition model improves the precision rate by 7%, the recall rate by 8%, and the average accuracy rate by 10%. The FPS of the model optimized by knowledge distillation is 62.5, and the average value of the FPS of other versions of this model is 47.35. It is believed that the two optimization ideas introduced in this paper can enhance the previous transmission line defect recognition algorithm in terms of accuracy and recognition speed.

In the field of light guide plate (LGP) quality inspection in industrial production, traditional object detection models often face various challenges, such as difficulties in extracting features of small defects, low accuracy in multi-scale defect detection, and interference from complex backgrounds. To effectively address these problems, this study focuses on exploring advanced object detection technologies and proposes the RT-DETR-LGP model. This model adopts the newly designed Multi-Scale Edge Information Enhancement (MSEIE) module and Aggregate Diffusion Pyramid Network (ADPN) module, and uses a multi-scale feature fusion strategy to achieve efficient detection of defects of different sizes. In addition, this study uniquely combines the model with Channel-Wise Knowledge Distillation (CWKD) technology to improve the model’s detection accuracy and generalization ability without increasing the number of model parameters. To comprehensively evaluate the performance of the RT-DETR-LGP model, 10831 LGP defect samples were carefully collected from industrial sites and used to create the ILGPDD dataset containing seven different types of defects. After integrating the knowledge distillation technology, the RT-DETR-LGP model demonstrated outstanding performance in all key indicators. The <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AP</i>, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AP</i><sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">50</sub>, and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AP</i><sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">75</sub> reached 70.3%, 98.0%, and 83.5% respectively, representing improvements of 2.5%, 1.5%, and 2.8% compared to the RT-DETR baseline network. Moreover, the model’s FPS reached 176, indicating its ability to achieve rapid LGP defect detection. These results indicate the great potential of the RT-DETR-LGP model in detecting surface defects of LGPs, providing an efficient and reliable solution for LGP quality inspection in industrial production.

Multispectral remote sensing image (MSI) semantic segmentation faces challenges of limited labeled data and significant scene variability. Although domain adaptation (DA) and domain generalization (DG) methods alleviate these issues to some extent, they still have limitations. DA requires target domain (TD) data, and DG has limited task adaptability. The recently emerged segment anything model (SAM) demonstrates exceptional zero-shot generalization capabilities, yet its visible-light training data and interactive prompt requirements prevent direct application to MSI segmentation tasks. To address these challenges, this article proposes a single-source domain defect-aware adaptation and style-modulated generalization network (SDSnet), which integrates two key innovations: defect-aware prompt learning that automatically focuses on high-difficulty regions through entropy-based defect detection, and style generalization learning that enhances cross-domain adaptability via codebook-based style modulation. Through knowledge distillation, SDSnet enables efficient inference using only the base network, without additional computational overhead. Extensive experiments on three TDs demonstrate SDSnet's superiority over state-of-the-art DA, DG, and SAM-based methods. Code will be available at https://github.com/zhaoboyu34526/SDSnet.

The design and study of pulsed eddy current sensors for detecting surface defects in small-diameter rods are highly significant. Accurate detection and identification of surface defects in small-diameter rods may be attained by the ongoing optimization of sensor design and enhancement of detection technologies. This article presents the construction of a non-coaxial differential eddy current sensor (Tx-Rx sensor) and examines the detection of surface defects in a small diameter bar. A COMSOL 3D model is developed to examine the variations in eddy current distribution and defect signal characteristics between the plate and rod components. The position of the excitation coil on the bar and the eddy current disruption around the defect are examined. Additionally, a Tx-Rx sensor has been developed and enhanced concerning coil dimensions, coil separation, and elevation height. An experimental system is established to detect bar structures with surface defects of varying depths, and a model correlating differential signal attenuation with defect depth is proposed, achieving a quantitative relative error of less than 5%, thereby offering a reference for the quantitative detection of bar surface defects.

Various methods of detecting surface defects are being used in automated industrial manufacturing environments. This work presents the design and development of a laser charged couple device (CCD) displacement scanning system. The surface defect detection method using a laser CCD displacement sensor is derived from an idea in which surface defects such as cracks, inclusions and holes of three‐dimensional morphology characteristics are compared to normal surface especially in continuous casting slabs. Some novel research methods have been applied to develop the surface defect detection system used: first, getting a one‐dimensional distance matrix along transverse direction and a two‐dimensional matrix combining with a certain moving speed through the laser CCD displacement sensor; second, obtaining the slab surface profile and mapping to a greyscale image and finally, obtaining the surface defect regions by an image processing and searching algorithm and quantitatively detecting slab surface defect shape and depth. The research results in the lab trials have shown that the methodology proposed is effective to detect two‐dimensional defect size and reconstruct a three‐dimensional surface defect shape. At the same time, it is also able to accurately locate and identify surface defects and realise automatic surface defect non‐destructive detection online, and provides a theoretical base and technology idea for further studying surface defects online inspection for hot continuous casting slab.

Carbon fiber reinforced polymer (CFRP) has become an essential structural material in aerospace, new energy, rail transportation, and other fields due to its lightweight and high-strength properties. However, its complex manufacturing processes and demanding service environments can lead to various types of damage on both the surface and inside the material, significantly compromising safety. Existing detection techniques often adopt a step-by-step inspection mode, which results in complex processes and makes it difficult to achieve cross-dimensional correlation analysis of defects, leading to blind spots in structural health assessment. To address these issues, this paper proposes a method for simultaneous detection of surface and subsurface defects in CFRP based on infrared polarization imaging. First, we developed an infrared polarization imaging-based defect detection system and collected a CFRP defect dataset of samples containing various surface and subsurface defects. Then, we proposed a surface and subsurface defect detection network, which integrates extracted thermal diffusion features and polarization features to decouple and enhance surface and subsurface features. Through the constraints of multi-task defect detection, simultaneous detection of surface and subsurface defects is achieved. The experiments on the collected CFRP defect dataset demonstrate the effectiveness of the proposed method.

Sanitary ceramic products, such as toilet and wash basin, are widely used in our daily life. Sanitary ceramics are expected to have some excellent physical properties, such as corrosion resistance, easy cleaning, and low water absorption. However, surface defects in sanitary ceramics are inevitable due to complex production processes and changing production environment. Therefore, surface defect detection must be performed in the manufacturing process of sanitary ceramics. There are many types of surface defects in sanitary ceramics, and different types of defects have large differences in characteristics and scales. Traditional detection methods with artificially designed features and classifiers are difficult to apply in this context. In addition, there are few studies on surface defect detection methods of sanitary ceramics based on deep neural networks. In this article, a lightweight real-time defect detection network based on the lightweight backbone MobileNetV3 is presented. The proposed network achieves multi-scale detection of surface defects in sanitary ceramics with a multi-layer feature pyramid. Combining region proposal network and anchor-free method, real-time defect detection is achieved. Finally, a detection head with channel attention structure and a low-level mixed feature classification strategy is used to perform defect classification with higher accuracy. Experimental results show that the proposed approach achieves at least 22.9&#x0025; detection speed improvement and 35.0&#x0025; average precision improvement while reducing memory consumption by at least 8.4&#x0025; compared with the classic one-stage SSD, YOLO V3 and two-stage Faster R-CNN methods.

Surface defect detection aims to classify and locate a certain defect that exists in the target surface area. It is an important part of industrial quality inspection. Most of the research on surface defect detection are currently based on convolutional neural networks (CNNs), which are more concerned with local information and lack global perception. Thus, CNNs are unable to effectively extract the defect features. In this paper, a defect detection method based on the Swin transformer is proposed. The structure of the Swin transformer has been fine-tuned so that it has five scales of output, making it more suitable for defect detection tasks with large variations in target size. A bi-directional feature pyramid network is used as the feature fusion part to efficiently fuse to the extracted features. The focal loss is used as a loss function to weight the hard- and easy-to-distinguish samples, potentially making the model fit the surface defect data better. To reduce the number of parameters in the model, a shared detection head was chosen for result prediction. Experiments were conducted on the flange surface defect dataset and the steel surface defect dataset, respectively. Compared with the classical CNNs target detection algorithm, our method improves the mean average precision (mAP) by about 15.4%, while the model volume and detection speed are essentially the same as those of the CNNs-based method. The experimental results show that our proposed method is more competitive compared with CNNs-based methods and has some generality for different types of defects.