Improved YOLOv7 Electric Work Safety Belt Hook Suspension State Recognition Algorithm Based on Decoupled Head

Coronary artery stenosis detection remains a challenging task due to the complex vascular structure, poor quality of imaging pictures, poor vessel contouring caused by breathing artifacts and stenotic lesions that often appear in a small region of the image. In order to improve the accuracy and efficiency of detection, a new deep-learning technique based on a coronary artery stenosis detection framework (DCA-YOLOv8) is proposed in this paper. The framework consists of a histogram equalization and canny edge detection preprocessing (HEC) enhancement module, a double coordinate attention (DCA) feature extraction module and an output module that combines a newly designed loss function, named adaptive inner-CIoU (AICI). This new framework is called DCA-YOLOv8. The experimental results show that the DCA-YOLOv8 framework performs better than existing object detection algorithms in coronary artery stenosis detection, with precision, recall, F1-score and mean average precision (mAP) at 96.62%, 95.06%, 95.83% and 97.6%, respectively. In addition, the framework performs better in the classification task, with accuracy at 93.2%, precision at 92.94%, recall at 93.5% and F1-score at 93.22%. Despite the limitations of data volume and labeled data, the proposed framework is valuable in applications for assisting the cardiac team in making decisions by using coronary angiography results.

Efficient feature representation is the key to improving crowd counting performance. CNN and Transformer are the two commonly used feature extraction frameworks in the field of crowd counting. CNN excels at hierarchically extracting local features to obtain a multi-scale feature representation of the image, but it struggles with capturing global features. Transformer, on the other hand, could capture global feature representation by utilizing cascaded self-attention to capture remote dependency relationships, but it often overlooks local detail information. Therefore, relying solely on CNN or Transformer for crowd counting has certain limitations. In this paper, we propose the TCHNet crowd counting model by combining the CNN and Transformer frameworks. The model employs the CMT (CNNs Meet Vision Transformers) backbone network as the Feature Extraction Module (FEM) to hierarchically extract local and global features of the crowd using a combination of convolution and self-attention mechanisms. To obtain more comprehensive spatial local information, an improved Progressive Multi-scale Learning Process (PMLP) is introduced into the FEM, guiding the network to learn at different granularity levels. The features from these three different granularity levels are then fed into the Multi-scale Feature Aggregation Module (MFAM) for fusion. Finally, a Multi-Scale Regression Module (MSRM) is designed to handle the multi-scale fused features, resulting in crowd features rich in high-level semantics and low-level detail. Experimental results on five benchmark datasets demonstrate that TCHNet achieves highly competitive performance compared to some popular crowd counting methods.

The accuracy of road crack detection is susceptible to factors such as illumination variations and tiny crack sizes. Additionally, the loss function in road crack detection is prone to fluctuations during the training process, leading to instability. To address these issues, this paper proposes a multimodal crack detection network to improve detection robustness and accuracy. First, a feature extraction module, named channel–spatial Kolmogorov–Arnold network (CSKAN), is designed specifically to handle challenges posed by illumination variations, enhancing the stability of feature extraction. (KAN is a mathematical framework that decomposes complex multivariate functions into learnable univariate components, enabling adaptive nonlinear modeling.) Second, an efficient multi-scale attention mechanism (Laplacian-adaptive weighted attention) based on the Laplace operator is introduced to address detection difficulties caused by tiny crack sizes and to improve the network's ability to capture small-scale crack features. Finally, a decoupling method for physical information neural networks (DPINNs) is proposed for the first time, ensuring better smoothness in the loss function during the training process. Experimental results demonstrate that our proposed detection framework outperforms seven state-of-the-art detection methods.

Many deep learning (DL)-based detectors have been developed for optical remote sensing object detection in recent years. However, most of the recent detectors are developed toward the pursuit of a higher accuracy, but little toward a balance between accuracy, deployability and inference time, which hinders the practical application for these detectors, especially in embedded devices. In order to achieve a higher detection accuracy and reduce the computational consumption and inference time simultaneously, a novel convolutional network named YOLO-DSD was developed based on YOLOv4. Firstly, a new feature extraction module, a dense residual (DenseRes) block, was proposed in a backbone network by utilizing a series-connected residual structure with the same topology for improving feature extraction while reducing the computational consumption and inference time. Secondly, convolution layer–batch normalization layer–leaky ReLu (CBL) ×5 modules in the neck, named S-CBL×5, were improved with a short-cut connection in order to mitigate feature loss. Finally, a low-cost novel attention mechanism called a dual channel attention (DCA) block was introduced to each S-CBL×5 for a better representation of features. The experimental results in the DIOR dataset indicate that YOLO-DSD outperforms YOLOv4 by increasing mAP0.5 from 71.3% to 73.0%, with a 23.9% and 29.7% reduction in Params and Flops, respectively, but a 50.2% improvement in FPS. In the RSOD dataset, the mAP0.5 of YOLO-DSD is increased from 90.0~94.0% to 92.6~95.5% under different input sizes. Compared with the SOTA detectors, YOLO-DSD achieves a better balance between the accuracy, deployability and inference time.

Face detection technology is widely used in various fields, such as security systems and authentication systems. In the current stage of normalized epidemic prevention and control, people need to wear masks in public. A new approach is urgently needed to enable face detection when people wear masks. First, according to the public dataset Widerface and MAFA, this paper uses a Dlib-based way to obtain a self-built dataset and label them. Second, we have made the following innovations to the YOLOv3 face detection algorithm. The original Darknet-53 is replaced with ResNet50 as the backbone network, where DCN(Deformable Convolutional Network) enhances the robustness of the model to slightly deformed targets. At the same time, the FPN(Feature Pyramid Network) is replaced by BiFPN(Bidirectional Feature Network) and DIoU(Distance IoU) is introduced as a positioning loss function. Third, after changing faceNet's loss function to ArcFace, two models are trained using FaceNet: a mask-wearing face and a normal face. Experiments show that the face detection model proposed in this paper has an mAP50(mean Average Precision) of 94.68%, which is 2.53% higher than the original YOLOv3 model. Through multi-model with different weights selected by self-determination, the face recognition can maximize the use of the current facial feature points and achieve a more superior face identity recognition effect.

To address the problems of low accuracy and poor robustness in vehicle small object detection for autonomous driving tasks, this study aims to propose an improved vehicle small object detection algorithm model based on YOLOv5. Firstly, some convolutions in the backbone network are replaced with receptive field attention convolutions, and the weights of the convolution kernels are dynamically assigned based on the importance of image features to ensure the extraction of important features. Secondly, adding a channel attention mechanism to the backbone network enhances the attention to small target features. Finally, the Focal-EIoU loss function is introduced to increase the attention on high-quality samples in the regression stage of object detection boxes. When the model is applied to the small object test set of the KITTI dataset, the precision rate, recall rate and mean average precision are 88.5%, 82.8%, and 84.9%, respectively, and the frame processing rate reaches 87.83FPS.

Volcanism is an important geological evolutionary process on the Moon. The study of lunar volcanic features is of great significance and value to understanding the geological evolution of the Moon better. Lunar domes are one of the essential volcanic features of the Moon. However, the existing lunar dome detection methods are still traditional manual or semiautomatic identification approaches that require extensive prior knowledge and have a complex identification process. Therefore, this paper proposes an automatic detection method based on improved YOLOv7 for lunar dome detection. First, a new lunar dome dataset was created by digital elevation model (DEM) data, and the effective squeeze and excitation (ESE) attention mechanism module was added to the backbone and neck sections to reduce information loss in the feature map and enhance network expressiveness. Then, a new SPPCSPC-RFE module was proposed by adding the receptive field enhancement (RFE) module into the neck section, which can adapt to dome feature maps of different shapes and sizes. Finally, the bounding box regression loss function complete IOU (CIOU) was replaced by wise IOU (WIOU). The WIOU loss function improved the model's performance for the dome detection effect. Furthermore, this study combined several data enhancement strategies to improve the robustness of the network. To evaluate the performance of the proposed model, we conducted several experiments using the dome dataset developed in this study. The experimental results indicate that the improved method outperforms related methods with a mean average precision (mAP@0.5) value of 88.7%, precision (P) value of 85.6%, and recall (R) value of 86.4%. This study provides an effective solution for lunar dome detection.

In response to the challenges in highway pavement distress detection, such as multiple defect categories, difficulties in feature extraction for different damage types, and slow identification speeds, this paper proposes an enhanced pavement crack detection model named Star-YOLO11. This improved algorithm modifies the YOLO11 architecture by substituting the original C3k2 backbone network with a Star-s50 feature extraction network. The enhanced structure adjusts the number of stacked layers in the StarBlock module to optimize detection accuracy and improve model efficiency. To enhance the accuracy of pavement crack detection and improve model efficiency, three key modifications to the YOLO11 architecture are proposed. Firstly, the original C3k2 backbone is replaced with a StarBlock-based structure, forming the Star-s50 feature extraction backbone network. This lightweight redesign reduces computational complexity while maintaining detection precision. Secondly, to address the inefficiency of the original Partial Self-attention (PSA) mechanism in capturing localized crack features, the convolutional prior-aware Channel Prior Convolutional Attention (CPCA) mechanism is integrated into the channel dimension, creating a hybrid CPC-C2PSA attention structure. Thirdly, the original neck structure is upgraded to a Star Multi-Branch Auxiliary Feature Pyramid Network (SMAFPN) based on the Multi-Branch Auxiliary Feature Pyramid Network architecture, which adaptively fuses high-level semantic and low-level spatial information through Star-s50 connections and C3k2 extraction blocks. Additionally, a composite dataset augmentation strategy combining traditional and advanced augmentation techniques is developed. This strategy is validated on a specialized pavement dataset containing five distinct crack categories for comprehensive training and evaluation. Experimental results indicate that the proposed Star-YOLO11 achieves an accuracy of 89.9% (3.5% higher than the baseline), a mean average precision (mAP) of 90.3% (+2.6%), and an F1-score of 85.8% (+0.5%), while reducing the model size by 18.8% and reaching a frame rate of 225.73 frames per second (FPS) for real-time detection. It shows potential for lightweight deployment in pavement crack detection tasks.

Sheep behavior recognition helps to monitor the health status of sheep and prevent the outbreak of infectious diseases. Aiming at the problems of low detection accuracy and slow speed due to the crowding of sheep in real farming scenarios, which can easily obscure each other, this study proposes a lightweight sheep behavior recognition model based on the YOLOv8n model. First, the Convolutional Block Attention Module (CBAM) is introduced and improved in the YOLOv8n model, and the channel attention module and spatial attention module are changed from serial to parallel to construct a novel attention mechanism, PCBAM, to enhance the network’s attention to the sheep and eliminate redundant background information; second, the ordinary convolution in the backbone network is replaced with depth-separable convolution, which effectively reduces the number of parameters in the model and reduces the computational complexity. The study takes the housed breeding sheep as the test object, installs a camera diagonally above the sheep pen to collect images and makes a data set for testing, and in order to verify the superiority of the PD-YOLO model, compares it with a variety of target detection models. The experimental results show that the mean average precision (mAP) of the model proposed in this paper are 95.8%, 98.9%, and 96.2% for the three postures of sheep lying, feeding, and standing, respectively, which are 8.5%, 0.8%, and 0.8% higher than those of YOLOv8n, respectively, and the size of the model has been reduced by 13.3% and the amount of computation has been reduced by 12.1%. The inference speed reaches 52.1 FPS per second, which is better than other models in meeting the real-time detection requirement. To verify the practicality of this research method, the PD-YOLO model was deployed on the RK3399Pro development board for testing, and a high inference speed was achieved. It can provide effective technical support for sheep smart farming.

A model of speech recognition system is proposed and simulated. This model consists of two modules: a feature extraction module which is composed of layers of neuron-like cells with time constants, and a recognition module with delay-lines. This paper offers a detailed explanation of the feature extraction module of this model. The module consists of an adaptive-Q filter bank, which corresponds to the basilar membrane, a halfwave rectifier block, which corresponds to the layer of hair cells, and a feature extraction block made of neural cells. The features extracted by this model are frequency ascending, frequency descending, and continuous constant frequency components.

Domestic trash detection is an essential technology toward achieving a smart city. Due to the complexity and variability of urban trash scenarios, the existing trash detection algorithms suffer from low detection rates and high false positives, as well as the general problem of slow speed in industrial applications. This paper proposes an i-YOLOX model for domestic trash detection based on deep learning algorithms. First, a large number of real-life trash images are collected into a new trash image dataset. Second, the lightweight operator involution is incorporated into the feature extraction structure of the algorithm, which allows the feature extraction layer to establish long-distance feature relationships and adaptively extract channel features. In addition, the ability of the model to distinguish similar trash features is strengthened by adding the convolutional block attention module (CBAM) to the enhanced feature extraction network. Finally, the design of the involution residual head structure in the detection head reduces the gradient disappearance and accelerates the convergence of the model loss values allowing the model to perform better classification and regression of the acquired feature layers. In this study, YOLOX-S is chosen as the baseline for each enhancement experiment. The experimental results show that compared with the baseline algorithm, the mean average precision (mAP) of i-YOLOX is improved by 1.47%, the number of parameters is reduced by 23.3%, and the FPS is improved by 40.4%. In practical applications, this improved model achieves accurate recognition of trash in natural scenes, which further validates the generalization performance of i-YOLOX and provides a reference for future domestic trash detection research.

Surface quality assessment of magnetic tile before mounting is extremely significant. At present, this task is mainly accomplished by experienced workers in industry, which exposes the drawbacks of low efficiency and high cost. To overcome these issues, an intelligent system is developed to perform this task, which appears to be an efficient and reliable substitute for human workers. In this article, deep learning technique is embedded into our system for automatic defect identification. However, conventional convolutional neural network (CNN) is not suitable for this classification task, since the input is a sample rather than a single image. To overcome this problem, an end-to-end CNN architecture is proposed, termed fusion feature CNN (FFCNN). FFCNN consists of three modules: feature extraction module, feature fusion module, and decision-making module. The feature extraction module is designed to extract features from different images. The feature fusion module is to fuse the features extracted by feature extraction module. The decision-making module is to predict the label by the fused features. Furthermore, an attention mechanism is introduced to focus on more representative parts and suppress less important information. Experimental results demonstrated that the developed system is effective and efficient for magnetic tile surface defect detection.

In response to the subjective issues, inconsistent quality standards, high labor intensity and low sorting efficiency during the drying process of green pepper, an improved YOLOv11 algorithm was proposed for quality detection. A multi-scale edge enhancement module (MEEM) is introduced into the backbone network, replacing the original basic C3K2 module with C3K2-MEEM to enhance the extraction of detailed features in images of dried green Sichuan pepper and prevent missed detections, false detections, and boundary confusion. The LRSA module is integrated into the 10th layer of the backbone network to improve the clarity of the tumor-like texture of the Sichuan pepper and reduce the influence of impurities, automatically allocating attention based on feature similarity to preserve local information. In the neck layer, the DPCF module is added to the FPN+PAN feature fusion stage to achieve multi-scale feature collaboration, meeting the detection requirements of dried green Sichuan pepper. The results show that the accuracy recall rate, mean average precision, and model size of the improved MLD-YOLOv11 algorithm are 92.1%, 96.6%, 95.6%, and 11.06 MB, respectively. Compared with the training results of the original YOLOv11 model, the average accuracy of the improved model has increased by 2.2 percentage points, and GFLOPs have definitely decreased by 2 G, with parameter reduction of approximately 3.10%. Compared with other mainstream models, the MLD-YOLOv11 model has significant advantages in terms of mean average precision, model size, and floating point operations per second, making it more suitable for industrial applications and providing an efficient, accurate, and lightweight solution for the quality detection of dried green Sichuan pepper.

Advanced Siamese visual object tracking architectures are jointly trained using pair-wise input images to perform target classification and bounding box regression. They have achieved promising results in recent benchmarks and competitions. However, the existing methods suffer from two limitations: First, though the Siamese structure can estimate the target state in an instance frame, provided the target appearance does not deviate too much from the template, the detection of the target in an image cannot be guaranteed in the presence of severe appearance variations. Second, despite the classification and regression tasks sharing the same output from the backbone network, their specific modules and loss functions are invariably designed independently, without promoting any interaction. Yet, in a general tracking task, the centre classification and bounding box regression tasks are collaboratively working to estimate the final target location. To address the above issues, it is essential to perform target-agnostic detection so as to promote cross-task interactions in a Siamese-based tracking framework. In this work, we endow a novel network with a target-agnostic object detection module to complement the direct target inference, and to avoid or minimise the misalignment of the key cues of potential template-instance matches. To unify the multi-task learning formulation, we develop a cross-task interaction module to ensure consistent supervision of the classification and regression branches, improving the synergy of different branches. To eliminate potential inconsistencies that may arise within a multi-task architecture, we assign adaptive labels, rather than fixed hard labels, to supervise the network training more effectively. The experimental results obtained on several benchmarks, i.e., OTB100, UAV123, VOT2018, VOT2019, and LaSOT, demonstrate the effectiveness of the advanced target detection module, as well as the cross-task interaction, exhibiting superior tracking performance as compared with the state-of-the-art tracking methods.

Multi-temporal polarimertic SAR is considered to be very effective in crop classification and cultivated land detection, which has received much attention from researchers. Currently, for most multi-temporal polarimetric SAR data classification methods, the simultaneous temporal-polarimetric-spatial feature extraction capability has not been exploited sufficiently. Also, the diversity of different time and different polarimetric features has not been taken into account sufficiently. In this paper, we propose a classification model that combines a dual-stream network as a temporal-polarimetric-spatial feature extraction module with Vision Transformer(ViT) called Temporal-Polarimetric-Spatial Transformer(TSPT) to address the above problems. Secondly, a 3 dimension(3D) convolutional attention module that enables the network to weight the temporal dimension, polarimetric feature dimension and spatial dimension is developed, according to their importance. Experimental results on both UAVSAR and RADARSAT-2 datasets show that the proposed method outperforms ResNet.