Search

For pedestrian detection in all weather conditions, multispectral imagery is the preferred solution for multimodal data acquisition. Due to the complementarity of multispectral data, the performance of pedestrian detection has been continuously improved. However, it is precise because of the diversity of data that the complexity of the model is increased. How to obtain a simple and lightweight high-performance network is the primary problem that the industry needs to solve. To solve this problem, this paper firstly proposes a lightweight high-performance network multi-receptive field and attention mechanism for multispectral pedestrian detection (MAPD). MAPD is a new scalable network based on multi-receptive field and attention mechanism. It cleverly combines multi-receptive field module and convolutional block attention module (CBAM) attention module to obtain multi-receptive field and attention module (MA). The module can be easily embedded into other network structures. After that, we analyze the proportion of pedestrian objects in the image to determine the receptive field range of the target, and use this range to design the multi-receptive field module of the network to obtain a network model suitable for detecting different object tasks. The MAPD network proposed in this paper has a very low parameter amount. Ablation studies on KAIST and CVC-14 datasets show that our method is effective and achieves state-of-the-art detection performance. [ABSTRACT FROM AUTHOR]

Occlusion or posture change of the face in natural scenes has typical asymmetry; however, an asymmetric face plays a key part in the lack of information available for facial expression recognition. To solve the problem of low accuracy of asymmetric facial expression recognition, this paper proposes a fusion of channel global features and a spatial local information expression recognition network called the "Channel–Spatial Integration Network" (CSINet). First, to extract the underlying detail information and deepen the network, the attention residual module with a redundant information filtering function is designed, and the backbone feature-extraction network is constituted by module stacking. Second, considering the loss of information in the local key area of face occlusion, the channel–spatial fusion structure is constructed, and the channel features and spatial features are combined to enhance the accuracy of occluded facial recognition. Finally, before the full connection layer, more local spatial information is embedded into the global channel information to capture the relationship between different channel–spatial targets, which improves the accuracy of feature expression. Experimental results on the natural scene facial expression data sets RAF-DB and FERPlus show that the recognition accuracies of the modeling approach proposed in this paper are 89.67% and 90.83%, which are 13.24% and 11.52% higher than that of the baseline network ResNet50, respectively. Compared with the latest facial expression recognition methods such as CVT, PACVT, etc., the method in this paper obtains better evaluation results of masked facial expression recognition, which provides certain theoretical and technical references for daily facial emotion analysis and human–computer interaction applications. [ABSTRACT FROM AUTHOR]

Multisource remote sensing images have rich features and high interpretability and are widely employed in many applications. However, highly unbalanced category distributions and complex backgrounds have created some difficulties in the application of remote sensing image semantic segmentation tasks, such as low accuracy of small target segmentation and inaccurate edge extraction. To solve these problems, in this paper, a feature map segmentation reconstruction module and an attention upsampling module are proposed. In the encoder part, the input feature map is equally segmented, and the segmented feature map is enlarged to effectively improve the small target feature information expression ability in the model. In the decoder part, the key segmentation and location information of shallow features are obtained using the global view. The deep semantic information and shallow spatial location information are fully combined to achieve a more refined upsampling operation. In addition, the attention mechanism of the spatial and channel squeeze and excitation block (scSE) is applied to pay more attention to important features and to suppress irrelevant background and redundant information. To verify the effectiveness of the proposed method, the WHU-OPT-SAR dataset and six state-of-the-art algorithms are utilized in comparative experiments. The experimental results show that our model has demonstrated the best performance and low computational complexity. With only approximately half the floating-point operation count and the number of model parameters of the MCANet model, which is specially designed for the dataset, our model surpasses MCANet by 1.52% and 1.53% in terms of mean intersection over union (mIoU) and F1 score, respectively. In particular, for small object regions such as roads and other categories, compared to the baseline model, the IoU and F1 score of our model are improved by 5.27% and 3.99% and by 5.68% and 5.65%, respectively. These results demonstrate the superior performance of our model in terms of accuracy and efficiency. [ABSTRACT FROM AUTHOR]

To solve the problems of complex surface texture of magnetic tile, cumbersome process of traditional detection algorithm and low detection accuracy, this paper proposes a deep learning-based detection model: RC-YOLOv5s. The model incorporates two new structures: Res-Head and Drop-CA, where Res-Head enhances the feature fusion and information exchange between different layer structures and Drop-CA alleviates the case that the model pays too much attention to the defect target and reduces the missed detection rate of the model. Compared with YOLOv5s, the detection accuracy of the proposed model is improved by 1.83%, the missing rate is reduced from 1.673 to 0.372%, and the average detection frame rate of a single image reaches 41.67, which meets the requirements of real-time and accuracy of tile surface detection. [ABSTRACT FROM AUTHOR]

Head Pose Estimation (HPE) has a wide range of applications in computer vision, but still faces challenges: (1) Existing studies commonly use Euler angles or quaternions as pose labels, which may lead to discontinuity problems. (2) HPE does not effectively address regression via rotated matrices. (3) There is a low recognition rate in complex scenes, high computational requirements, etc. This paper presents an improved unconstrained HPE model to address these challenges. First, a rotation matrix form is introduced to solve the problem of unclear rotation labels. Second, a continuous 6D rotation matrix representation is used for efficient and robust direct regression. The RepVGG-A2 lightweight framework is used for feature extraction, and by adding a multi-level feature fusion module and a coordinate attention mechanism with residual connection, to improve the network's ability to perceive contextual information and pay attention to features. The model's accuracy was further improved by replacing the network activation function and improving the loss function. Experiments on the BIWI dataset 7:3 dividing the training and test sets show that the average absolute error of HPE for the proposed network model is 2.41. Trained on the dataset 300W_LP and tested on the AFLW2000 and BIWI datasets, the average absolute errors of HPE of the proposed network model are 4.34 and 3.93. The experimental results demonstrate that the improved network has better HPE performance. [ABSTRACT FROM AUTHOR]

In order to solve the problems of poor anti-interference and invasion of user privacy in traditional methods of using cameras for human activity recognition, this paper proposed a human activity recognition network based on spatiotemporal attention of millimeter wave radar 3D point cloud data to achieve accurate perception of intelligent application context. The network firstly used a secondary sliding time window to accumulate and separate the point cloud data generated by human activities as the input of the classifier, then used the PointLSTM unit to aggregate point features and states according to the point cloud coordinate relationship to extract the time sequence features of human activities, and then spliced temporal-spatial features, reduced the overall network computation and enhanced the network’s aggregation ability for local featured through sampling grouping modules, and finally used a stacked attention module to deeply fuse global and local features in temporal-spatial point cloud data to complete the accurate classification of human activities. This paper used millimeter wave radar to collect point cloud datasets of various human activities, the experimental results show that the average accuracy of the proposed spatiotemporal attention network can reach 98.64%,which can effectively identify complex and small-difference human activities, and meet the requirements of the human activity recognition system. [ABSTRACT FROM AUTHOR]

This article proposes a CBAM-ASPP-SqueezeNet model based on the attention mechanism and atrous spatial pyramid pooling (CBAM-ASPP) to solve the problem of robot multi-target grasping detection. Firstly, the paper establishes and expends a multi-target grasping dataset, as well as introduces and uses transfer learning to conduct network pre-training on the single-target dataset and slightly modify the model parameters using the multi-target dataset. Secondly, the SqueezeNet model is optimized and improved using the attention mechanism and atrous spatial pyramid pooling module. The paper introduces the attention mechanism network to weight the transmitted feature map in the channel and spatial dimensions. It uses a variety of parallel operations of atrous convolution with different atrous rates to increase the size of the receptive field and preserve features from different ranges. Finally, the CBAM-ASPP-SqueezeNet algorithm is verified using the self-constructed, multi-target capture dataset. When the paper introduces transfer learning, the various indicators converge after training 20 epochs. In the physical grabbing experiment conducted by Kinova and SIASUN Arm, a network grabbing success rate of 93% was achieved. [ABSTRACT FROM AUTHOR]

The traditional behavior recognition model has the disadvantage that it can't get the internal relationship between similar behaviors, such as smoking, pen, chin and the clamped objects, which limits the actual landing of such fine and complex behaviors as smoking recognition. To solve these problems, this paper puts forward the heterogeneous algorithm HMMA-NET (Heterogeneous multi-task smoking behavior recognition model combined with Attention), which consists of two modules: behavior prior and local detection, aiming at establishing the relationship between behavior and behavior objects. CNN combined with channel attention mechanism is used in both behavior prior module and local detection module. The former uses sign language semantic features to complete the primary prior of behavior according to the obtained behavior affinity vector field, while the latter designs network optimization such as fast Edgebox to obtain candidate areas, so as to transfer component information and achieve the goal of fast fine-grained detection. Finally, the two modules use SaaS mode to complete association recognition. Experiment shows that the algorithm can recognize complex actions effectively, and its accuracy is still equal to or even better than that of a single model, in which the accuracy of detecting smoking behavior scenes is 96.10%, and the false detection rate is 3.6%. The algorithm has been commercialized and applied to the actual monitoring of petrochemical scenes. The running results show that the algorithm can maintain good real-time performance and generalization ability. [ABSTRACT FROM AUTHOR]

Object detection based on wood defects involves using bounding boxes to label defects in the surface image of the wood. This step is crucial before the transformation of wood products. Due to the small size and diverse shape of wood defects, most previous object detection models are unable to filter out critical features effectively. Consequently, they have faced challenges in generating adequate contextual information to detect defects accurately. In this paper, we proposed a YOLOv5 model based on a Semi-Global Network (SGN) to detect wood defects. Unlike previous models, firstly, a lightweight SGN is introduced in the backbone to model the global context, which can improve the accuracy and reduce the complexity of the network at the same time; the backbone is embedded with the Extended Efficient Layer Aggregation Network (E-ELAN), which continuously enhances the learning ability of the network; and finally, the Efficient Intersection and Merger (EIOU) loss is used to solve the problems of slow convergence speed and inaccurate regression results. Experimental results on public wood defect datasets demonstrated that our approach outperformed existing target detection models. The mAP value was 86.4%, a 3.1% improvement over the baseline network model, a 7.1% improvement over SSD, and a 13.6% improvement over Faster R-CNN. These results show the effectiveness of our proposed methodology. [ABSTRACT FROM AUTHOR]

Synthetic Aperture Radar (SAR) can provide rich feature information under all-weather and day-night conditions because it is not affected by climatic conditions. However, multiplicative speckle noise exists in SAR images, which makes it difficult to accurately identify some fuzzy targets in SAR images, such as roads and rivers, during semantic segmentation. This paper proposes an improved Deeplabv3+ network that can be effectively applied to the semantic segmentation task of SAR images. Firstly, this paper added the attention mechanism and, combined with the idea of an image pyramid, proposed the Feature Post-Processing Module (FPPM) to post-process the network output feature map, obtain better fine image features, and solve the problem of fuzzy texture and spectral features of SAR images. Compared to the original Deeplabv3+ network, the segmentation accuracy has been improved by 3.64% and mIoU improved by 1.09%. Secondly, to solve the problems of limited SAR image data and an unbalanced sample, this paper used the focal loss function to improve the backbone function of the network, which increased the mIoU by 1.01%. Finally, the Atrous Spatial Pyramid Pooling (ASPP) module was improved and the 3 × 3 void convolution in ASPP was decomposed into 2D, which can maintain the void ratio and effectively reduce the calculation amount of the module, shorten the training time by 19 ms and improve the semantic segmentation effect. [ABSTRACT FROM AUTHOR]

Local feature information plays an important role in image segmentation, however, the referring image segmentation task is dependent on the text expression, so it is impossible to extract local feature information directly from the original reference image. In order to solve this problem, this paper proposes a specific noun-guided local feature extraction deep neural network model（NgLFNet). The NgLFNet model can automatically mine the local feature information of the object to be segmented according to the key nouns in the input text expression. Specifically, the model first obtains key nouns in text through sentence analysis; secondly, extracting corresponding features through text and image encoders, and using the key nouns to obtain local features of high-interest regions through the multihead attention mechanism; then the multi-modal features are gradually fused to learn; Finally, the decoding and correction module uses the obtained local features to perform more detailed corrections on the prediction mask to obtain the final result. The method in this paper is compared with a variety of mainstream referring segmentation methods and the experimental results show that the proposed method improves the accuracy of text-based instance segmentation task. [ABSTRACT FROM AUTHOR]

In order to further mine the changelog information, this paper proposed a siamese Bi-lSTM network model based on the attention mechanism to classify the changelog to realize topic annotation and assist in locating the location of code defects. The model proposed a word segmentation tool with security features to realize changelog preprocessing, used Bi-lSTM network to learn changelog contextual semantic information, and solved the problem of overfilling mode existing in changelog itself through siamese neural network and expanded the data set with high quality to improve generalization ability. It carried out serialization training for the changelog composed of multiple sentences, and the influence of sentences was distinguished through the attention mechanism. For some changelogs of the defect repair class, this paper improved the LLVM tool, generated a mapping table to search the log content, and located the location of the defect module in the source code. A large number of experimental results show that the classification effect of the model in this paper has strong generalization ability, and is nearly 10% higher than the general methods in text classification methods in terms of accuracy, F, value and other indicators, has ideal log classification effect and topic learning effect. [ABSTRACT FROM AUTHOR]

The application of mulching film has significantly contributed to improving agricultural output and benefits, but residual film has caused severe impacts on agricultural production and the environment. In order to realize the accurate recycling of agricultural residual film, the detection of residual film is the first problem to be solved. The difference in color and texture between residual film and bare soil is not obvious, and residual film is of various sizes and morphologies. To solve these problems, the paper proposes a method for detecting residual film in agricultural fields that uses the attention mechanism. First, a two-stage pre-training approach with strengthened memory is proposed to enable the model to better understand the residual film features with limited data. Second, a multi-scale feature fusion module with adaptive weights is proposed to enhance the recognition of small targets of residual film by using attention. Finally, an inter-feature cross-attention mechanism that can realize full interaction between shallow and deep feature information to reduce the useless noise extracted from residual film images is designed. The experimental results on a self-made residual film dataset show that the improved model improves precision, recall, and mAP by 5.39%, 2.02%, and 3.95%, respectively, compared with the original model, and it also outperforms other recent detection models. The method provides strong technical support for accurately identifying farmland residual film and has the potential to be applied to mechanical equipment for the recycling of residual film. [ABSTRACT FROM AUTHOR]

Objectives: Sea-land segmentation is of great significance for tasks such as ocean target detection and coastline extraction in synthetic aperture radar (SAR) image. To solve the problem of sea-land segmentation of multi-resolution SAR image in practical applications, this paper presents a sea-land segmentation method based on context and edge attention. Methods: The proposed method uses the channel attention mechanism to fuse context features of different scales and levels, and designs an edge extraction branch to provide edge information for further improving the segmentation result of the boundary area. In addition, a sea-land segmentation dataset of multi-resolution SAR image based on the Gaofen-3 satellite data is provided. The dataset covers multiple resolution images, including various sea-land boundary types such as ports, islands. Using this multi-resolution SAR image coastline segmentation dataset, we perform experimental analyses on the effectiveness of the proposed network and the contributions of each module. Results and Conclusions: Experimental results show that the proposed method can work well for the task of sealand segmentation, the overall classification accuracy and mean intersection over union achieve 98.21% and 96.47%, respectively. [ABSTRACT FROM AUTHOR]

As an infrastructure of biochemical laboratories, EP tube label plays a significant role in information extraction to meet the limitations of computing power in offline devices and solve the problem that the EP tube label cannot be accurately identified before identification because the label belongs to multi-angle random placement. This paper proposes a light-weight neural network YOLOv4-tiny-ECA to position tubes and a tilt correction method based on Hough transform. First, the EP tube rack is roughly positioned based on the diffuse filling algorithm combined with digital morphological corrosion, and then the EP tubes in the rack are precisely positioned using the light-weight YOLO target detection algorithm combined with the attention mechanism. Next, the baseline is added to the label as the basis for determining the tilt angle. For the valid target, the baseline is extracted using the Hough transform and the tilt angle is calculated by vector fork multiplication. Finally, baseline is removed using image processing algorithm for better recognition results. Our results show that the light-weight YOLO algorithm reduces the network parameters by 56% and computation by 55% while keeping the accuracy rate largely unchanged, the offline positioning tilt correction method can achieve 98.8% accuracy and 0.076 s processing speed for a single test tube on average, which meets the real-time requirement. [ABSTRACT FROM AUTHOR]

With the rapid development of computer vision and artificial intelligence technology, visual object detection has made unprecedented progress, and small object detection in complex scenes has attracted more and more attention. To solve the problems of ambiguity, overlap and occlusion in small object detection in complex scenes. In this paper, a multi-scale fusion feature enhanced path aggregation network MSFE-PANet is proposed. By adding attention mechanism and feature fusion, the fusion of strong positioning information of deep feature map and strong semantic information of shallow feature map is enhanced, which helps the network to find interesting areas in complex scenes and improve its sensitivity to small objects. The rejection loss function and network prediction scale are designed to solve the problems of missing detection and false detection of overlapping and blocking small objects in complex backgrounds. The proposed method achieves an accuracy of 40.7% on the VisDrone2021 dataset and 89.7% on the PASCAL VOC dataset. Comparative analysis with mainstream object detection algorithms proves the superiority of this method in detecting small objects in complex scenes. [ABSTRACT FROM AUTHOR]

In this paper, we propose a lightweight asymmetric refining fusion network (LARFNet) for real-time semantic segmentation to solve the problem that some existing models cannot achieve good segmentation accuracy with real-time inference speed in mobile devices due to the huge computational overhead. Specifically, LARFNet adopts an asymmetric encoder–decoder structure. The depth-wise separable asymmetric interaction module (DSAI module) is designed in the encoding process, which effectively extracted local and surrounding information under different receptive fields with optimized convolution in the condition of ensuring communication between channels. In the decoder, we design the bilateral pyramid pooling attention module (BPPA module) and the multi-stage refinement fusion module (MRF Module). The BPPA module is used to integrate the high-level output multi-scale context information. Based on spatial and channel attention mechanisms, the MRF module is proposed to refine the feature maps of different resolutions and guide the feature fusion. Experimental results show that LARFNet achieves 69.2% mIoU and 65.6% mIoU on Cityscapes and Camvid datasets at 127 FPS and 222 FPS respectively, only using a single NVIDIA GeForce GTX2080Ti GPU and 0.72M parameters without any pre-training or pre-processing. Compared with most of the existing state-of-the-art models, the proposed method realizes the efficient use of network parameters at a faster speed, reduces the number of network parameters, and still achieves the accuracy of good segmentation. [Display omitted] • In this paper, we propose a lightweight real-time semantic segmentation network which considers inference speed, number of model parameters and segmentation accuracy as a whole, named lightweight asymmetric refining fusion network (LARFNet). The network mainly consists of three modules. The DSAI module is used to extract different features. The BPPA module is used to provide pixel-level attention for features, and the MRF module is used to guide feature fusion after optimizing features. [ABSTRACT FROM AUTHOR]

Small object detection has always been a difficult direction in the field of object detection, especially the detection of small objects in UAV aerial images. The images captured by UAVs have the characteristics of small objects and dense objects. In order to solve these two problems, this paper improves the performance of object detection from the aspects of data and network structure. In terms of data, the data augmentation strategy and image pyramid mechanism are mainly used. The data augmentation strategy adopts the method of image division, which can greatly increase the number of small objects, making it easier for the algorithm to be fully trained during the training process. Since the object is denser, the image pyramid mechanism is used. During the training process, the divided images are up-sampled into three different sizes, and then sent to three different detectors respectively. Finally, the detection results of the three detectors are fused to obtain the final detection results. The small object itself has few pixels and few features. In order to improve the detection performance, it is necessary to use context. This paper adds attention mechanism to the yolov5 network structure, while adding a detection head to the underlying feature map to make the network structure pay more attention to small objects. By using data augmentation and improved network structure, the detection performance of small objects can be significantly improved. The experiment in this paper is carried out on the Visdrone2019 dataset and DOTA dataset. Through experimental verification, our proposed method can significantly improve the performance of small object detection. [ABSTRACT FROM AUTHOR]

Due to the heterogeneity of program language and natural language query, it is difficult to identify the semantic relationship between them, which leads to the low efficiency of code recommendation. In order to solve the problems of the above code recommendation technology, a code recommendation method based on joint embedded attention network (JEAN) is proposed in this paper. The method uses GRU Network to embed code snippets and describe queries into vector representation, which solves the problem of heterogeneous code snippets and natural language queries. The Attention mechanism is then used to distribute totally different weights to different components of every mode of the code snippet. The reason for the Attention mechanism is that different components of every mode of the code snippet contribute differently to the semantic vector of the final code snippet, making it interpretable. Finally, two commonly used evaluation indexes of information retrieval, SuccessRate@k and MRR, are used for experimental comparison with other baseline models. The experimental results show that the code recommendation method based on joint embedded attention network proposed in this paper can effectively recommend appropriate code snippets according to the needs of developers, and its performance is better than other baseline methods. [ABSTRACT FROM AUTHOR]

In order to solve the problems of large number of parameters and high GPU resource cost in binocular vision stereo matching method based on neural network in robot and unmanned driving fields, this paper proposed a multi-scale aggregation stereo matching method. Firstly, this paper proposed a multi-scale feature extraction network to obtain richer features without changing the resolution by using dilated convolution, and introduced the attention mechanism was. Then, features at different resolutions were cross-fused to improve the feature information. Secondly, the acquisition method of cost volume was changed, the cost volume was obtained by aggregation at low scale, and continuously combined the high-scale similar information to update iteratively, and cross fused multiple cost volumes to obtain the final cost volume. Finally, combined with the refinement module of attention mechanism, the outliers and discontinuous regions in the initial disparity map are corrected to obtain the final disparity map. Experimental results show that the algorithm can run under low parameter number and low cost GPU resources, and obtain good matching accuracy. [ABSTRACT FROM AUTHOR]

Most existing text classification algorithms based on graph neural network ignore the problem of over-smoothing, and ignore the problem of information loss due to graph topology, resulting in poor classification performance. To solve this problem, this paper proposed a method to measure the smoothness of multiple text graph representations WACD and a regularization term RWACD to suppress over-smoothing. Subsequently, this paper proposed an attention and residual-based network structure ARS to compensate for the loss of textual information due to graph topology differences. Finally, this paper proposed a graph convolutional neural network text classification algorithm RA-GCN. RA-GCN used ARS to fuse text representations in the graph representation learning layer, and used RWACD in the readout layer to suppress over-smoothing. This paper conducted experiments on 6 Chinese and English datasets. The experimental results demonstrate the classification performance of RA-GCN, and the effects of RWACD and ARS are verified through multiple comparative experiments. [ABSTRACT FROM AUTHOR]

Most deep-learning-based object detection algorithms exhibit low speeds and accuracy in gear surface defect detection due to their high computational costs and complex structures. To solve this problem, a lightweight model for gear surface defect detection, namely STMS-YOLOv5, is proposed in this paper. Firstly, the ShuffleNetv2 module is employed as the backbone to reduce the giga floating-point operations per second and the number of parameters. Secondly, transposed convolution upsampling is used to enhance the learning capability of the network. Thirdly, the max efficient channel attention mechanism is embedded in the neck to compensate for the accuracy loss caused by the lightweight backbone. Finally, the SIOU_Loss is adopted as the bounding box regression loss function in the prediction part to speed up the model convergence. Experiments show that STMS-YOLOv5 achieves frames per second of 130.4 and 133.5 on the gear and NEU-DET steel surface defect datasets, respectively. The number of parameters and GFLOPs are reduced by 44.4% and 50.31%, respectively, while the mAP@0.5 reaches 98.6% and 73.5%, respectively. Extensive ablation and comparative experiments validate the effectiveness and generalization capability of the model in industrial defect detection. [ABSTRACT FROM AUTHOR]

The purpose of knowledge-based question answering (KBQA) is to accurately answer the questions raised by users through knowledge triples. Traditional Chinese KBQA methods rely heavily on artificial features, resulting in unsatisfactory QA results. To solve the above problems, this paper divides Chinese KBQA into two parts: entity extraction and attribute mapping. In the entity extraction stage, the improved Bi-LSTM-CNN-CRF model is used to identify the entity of questions and the Levenshtein distance method is used to resolve the entity link error. In the attribute mapping stage, according to the characteristics of questions and candidate attributes, the MGBA-LSTM-CNN model is proposed to encode questions and candidate attributes from the semantic level and word level, respectively, and splice them into new semantic vectors. Finally, the cosine distance is used to measure the similarity of the two vectors to find candidate attributes most similar to questions. The experimental results show that the system achieves good results in the Chinese question and answer data set. [ABSTRACT FROM AUTHOR]

In order to solve the problem of reducing the accuracy of part segmentation due to the loss of secondary features by PointNet max pooling, this paper proposed attention weighted feature aggregation Point Net network for part segmentation, and it could make full use of different features of point cloud for part segmentation. Firstly, it used the multiple layer perceptions to extract the spatial geometric features of the point cloud, and then multiplied the features and the shared weight matrix to obtain the attention scores of each feature of each point. Then it used the normalized scores as the weight to multiply the corresponding global feature and summed to obtain the aggregated global feature. Finally, it used the multiple layer perceptions to map the aggregated feature to the part segmentation results. Experimental results show that the proposed method improves the overall average mean intersection over union of part segmentation than traditional PointNet, has significant advantages in network robustness and computational complexity, and effectively optimizes PointNet. [ABSTRACT FROM AUTHOR]

There is a huge demand for weather cloud images in the observatory' s weather monitoring system. In order to solve the problems of model instability and loss of image features when the conventional generative adversarial network expands the dataset of the weather cloud images, this paper proposed a double-layer embedded adversarial image generation method based on SAU-Net DCGAN. This method consisted of two layers of networks which were nested within each other. Firstly,by the first layer of embedded network, it added the U-shaped network to the generator of the generative adversarial network. This network acted as the basic architecture and enhanced the feature recovery capability of the image by using the jump connection between the encoder and the decoder. Secondly,by the second layer of embedded network, it added SA to the U-shaped network. This attention mechanism reduced the complexity of the model by simplifying the parameters, improved effectively the feature loss in the dark part of the image. Finally, it developed a new weight calculation method to strengthen the connection between each features and improved the extraction of detail texture features from the images. The experimental results show that the quality of the images generated by this method is better than that of the conventional generative adversarial network in terms of sharpness and saturation. The evaluation indicators PSNR and SSIM have increased by 27. 06 dB and 0. 606 5 respectively. [ABSTRACT FROM AUTHOR]

For image denoising, this paper proposed CEANet which had a dual-channel dilated convolution with attention mechanism to solve the problem of information loss caused by deep neural network. Reserving block merged output feature maps of each layer to make up the loss of detail information during convolution. Dilated convolution achieved better balance between denoising performance and efficiency,extracting more features with less parameters and enhancing the representation capability of the model for noisy images. The sparse module of dilated convolution expanded the receptive field to extract significant structural information and edge features and recover details of complicated noisy images. The feature enhancement module based on attention mechanism further guided network for image denoising by fusing global features with local features. The experimental results show that CEANet achieves high peak signal-to-noise ratio and structure similarity mean value at Gaussian white noise level of 25 and 50, which can capture image detail information more efficiently and has better performance in edge retention and noise suppression. Through the above comparative experiments proved the effectiveness of the algorithm framework. [ABSTRACT FROM AUTHOR]

Target detection at the hive gate of a beehive can be used to effectively monitor invasive beehive species. However, in the natural environment, there is often a multi-target and multi-scale problem at the hive gate, making it difficult for beekeepers to accurately detect the internal state of the hive. (1) To solve the above problems, this paper proposes an improved RetinaNet target detection network, DY-RetinaNet, for the identification of common species at the hive doors of beehives in natural environments, i.e., Chinese bees, wasps, and cockroaches. (2) First, to solve the multi-target multi-scale problem presented in this paper, we propose replacing the FPN layer in the initial model RetinaNet with a symmetric structure BiFPN layer consisting of a feature pyramid, which allows the model to better balance the feature information of different scales. Then, for the loss function, using CIOU loss instead of smooth L1 loss makes the network more accurate for small target localization at multiple scales. Finally, the dynamic head framework is added after the model backbone network, due to the benefits of its multi-attention mechanism, which makes the model more concerned with multi-scale recognition in a multi-target scenario. (3) The experimental results of the homemade dataset show that DY-RetinaNet has the best network performance, compared to the initial model RetinaNet, when the backbone network is ResNet-101-BiFPN, and the mAP value of DY-RetinaNet is 97.38%. Compared with the initial model, the accuracy is improved by 6.77%. The experimental results from the public dataset MSCOCO 2017 show that DY-RetinaNet is better than the existing commonly used target-detection algorithms, such as SSD, YOLOV3, Faster R-CNN, Mask R-CNN, FCOS, and ExtremeNet. These results verify that the model has strong recognition accuracy and generalization ability for multi-target multi-scale detection. [ABSTRACT FROM AUTHOR]

Fast and accurate smoke detection is very important for reducing fire damage. Due to the complexity and changeable nature of smoke scenes, existing smoke detection technology has the problems of a low detection rate and a high false negative rate, and the robustness and generalization ability of the algorithms are not high. Therefore, this paper proposes a smoke detection model based on the improved YOLOv5. First, a large number of real smoke and synthetic smoke images were collected to form a dataset. Different loss functions (GIoU, DIoU, CIoU) were used on three different models of YOLOv5 (YOLOv5s, YOLOv5m, YOLOv5l), and YOLOv5m was used as the baseline model. Then, because of the problem of small numbers of smoke training samples, the mosaic enhancement method was used to randomly crop, scale and arrange nine images to form new images. To solve the problem of inaccurate anchor box prior information in YOLOv5, a dynamic anchor box mechanism is proposed. An anchor box was generated for the training dataset through the k-means++ clustering algorithm. The dynamic anchor box module was added to the model, and the size and position of the anchor box were dynamically updated in the network training process. Aiming at the problem of unbalanced feature maps in different scales of YOLOv5, an attention mechanism is proposed to improve the network detection performance by adding channel attention and spatial attention to the original network structure. Compared with the traditional deep learning algorithm, the detection performance of the improved algorithm in this paper was is 4.4% higher than the mAP of the baseline model, and the detection speed reached 85 FPS, which is obviously better and can meet engineering application requirements. [ABSTRACT FROM AUTHOR]

Due to the lack of sufficient co-occurrence information in short text, weak connections between words, and it is difficult to obtain subject words, which leads to the need to manually label a large number of training samples for short text classification, and the problems of sparse features and dimension explosion. This paper proposed a word symbiotic short text classification model based on attention mechanism and label graph（WGA-BERT）.Firstly, this paper used the pretrained BERT model to calculate the context aware text representation, and used WNTM to model the potential word group distribution of each word to obtain the topic expansion feature vector. Secondly, this paper used a tag graph construction method to capture the structure and relevance of subject words. Finally, this paper used an attention mechanism to establish the relationship between subject words and between subject words and text, which solved the problems of data sparsity and subject text heterogeneity. The experimental results show that the WGA-BERT model improves the classification accuracy by an average of 3% compared with the traditional machine learning model. [ABSTRACT FROM AUTHOR]

The grade of wheat quality depends on the proportion of unsound kernels. Therefore, the rapid detection of unsound wheat kernels is important for wheat rating and evaluation. However, in practice, unsound kernels are hand-picked, which makes the process time-consuming and inefficient. Meanwhile, methods based on traditional image processing cannot divide adherent particles well. To solve the above problems, this paper proposed an unsound wheat kernel recognition algorithm based on an improved mask RCNN. First, we changed the feature pyramid network (FPN) to a bottom-up pyramid network to strengthen the low-level information. Then, an attention mechanism (AM) module was added between the feature extraction network and the pyramid network to improve the detection accuracy for small targets. Finally, the regional proposal network (RPN) was optimized to improve the prediction performance. Experiments showed that the improved mask RCNN algorithm could identify the unsound kernels more quickly and accurately while handling adhesion problems well. The precision and recall were 86% and 91%, respectively, and the inference time on the test set with about 200 targets for each image was 7.83 s. Additionally, we compared the improved model with other existing segmentation models, and experiments showed that our model achieved higher accuracy and performance than the other models, laying the foundation for wheat grading. [ABSTRACT FROM AUTHOR]

On-site management of construction sites has always been a significant problem faced by the construction industry. With the development of UAVs, their use to monitor construction safety and progress will make construction more intelligent. This paper proposes a multi-category target detection system based on UAV low-altitude remote sensing, aiming to solve the problems of relying on fixed-position cameras and a single category of established detection targets when mainstream target detection algorithms are applied to construction supervision. The experimental results show that the proposed method can accurately and efficiently detect 15 types of construction site targets. In terms of performance, the proposed method achieves the highest accuracy in each category compared to other networks, with a mean average precision (mAP) of 82.48%. Additionally, by applying it to the actual construction site, the proposed system is confirmed to have comprehensive detection capability and robustness. [ABSTRACT FROM AUTHOR]

Relation detection for knowledge base question answering (KBQA) aims to select the best relation path that matches the question expressed by natural language from the candidate relations in the knowledge base and retrieve the answer to the question. To solve the problem of semantic information loss and inadequate attention interaction in existing relation detection methods, this paper proposed a multi-granularity relation detection model incorporating global-local features. The model firstly used bi-directional long short-term memory (Bi-LSTM) networks to encode the question and relation and modeled relations from multiple granularity, such as word-level and relation-level representation. Then, the model introduced a bi-directional attention mechanism to implement attentive interaction of the question and relation. Finally, it extracted global and local features by the aggregation operation and word-level interaction, respectively, to calculate the semantic similarity of the question and candidate relation. Experiments show that the proposed model achieves 93.5 % and 84.13% accuracy on SimpleQuestions and WebQuestionsSP datasets, respectively, improving the effect of relation detection. [ABSTRACT FROM AUTHOR]

The recommendation algorithm based on graph neural network can extract the association relationship between users and goods. Traditional methods can’t extract this relationship. At present, most of these algorithms ignore the general prefe-rences in the review data of users and products. In order to solve this problem, this paper proposed a new method. This me-thod used the graph neural network to extract association relations, and the advantage of deep learning to extract the general preferences, and carried out feature fusion to improve the recommendation effect. This paper conducted comparative experiments and ablation experiments on four sets of public data sets to verify the effectiveness of the proposed method. The evaluation indexes include the recall rate and normalized discounted cumulative gain. Experiments show that this method is more effective than the existing algorithms. The feature fusion of the two networks can improve the recommendation effect. [ABSTRACT FROM AUTHOR]

Logo detection is one of the crucial branches in computer vision due to various real-world applications, such as automatic logo detection and recognition, intelligent transportation, and trademark infringement detection. Compared with traditional handcrafted-feature-based methods, deep learning-based convolutional neural networks (CNNs) can learn both low-level and high-level image features. Recent decades have witnessed the great feature representation capabilities of deep CNNs and their variants, which have been very good at discovering intricate structures in high-dimensional data and are thereby applicable to many domains including logo detection. However, logo detection remains challenging, as existing detection methods cannot solve well the problems of a multiscale and large aspect ratios. In this paper, we tackle these challenges by developing a novel long-range dependence involutional network (LDI-Net). Specifically, we designed a strategy that combines a new operator and a self-attention mechanism via rethinking the intrinsic principle of convolution called long-range dependence involution (LD involution) to alleviate the detection difficulties caused by large aspect ratios. We also introduce a multilevel representation neural architecture search (MRNAS) to detect multiscale logo objects by constructing a novel multipath topology. In addition, we implemented an adaptive RoI pooling module (ARM) to improve detection efficiency by addressing the problem of logo deformation. Comprehensive experiments on four benchmark logo datasets demonstrate the effectiveness and efficiency of the proposed approach. [ABSTRACT FROM AUTHOR]

Semantic segmentation of remote sensing images plays a crucial role in urban planning and development. How to perform automatic, fast, and effective semantic segmentation of considerable size and high-resolution remote sensing images has become the key to research. However, the existing segmentation methods based on deep learning are complex and often difficult to apply practically due to the high computational cost of the excessive parameters. In this paper, we propose an end-to-end lightweight progressive attention semantic segmentation network (LPASS-Net), which aims to solve the problem of reducing computational costs without losing accuracy. Firstly, its backbone features are based on a lightweight network, MobileNetv3, and a feature fusion network composed of a reverse progressive attentional feature fusion network work. Additionally, a lightweight non-local convolutional attention network (LNCA-Net) is proposed to effectively integrate global information of attention mechanisms in the spatial dimension. Secondly, an edge padding cut prediction (EPCP) method is proposed to solve the problem of splicing traces in the prediction results. Finally, evaluated on the public datasets BDCI 2017 and ISPRS Potsdam, the mIoU reaches 83.17% and 88.86%, respectively, with an inference time of 0.0271 s. [ABSTRACT FROM AUTHOR]

In the field of CXR image classification, current studies mostly use convolutional neural network and transfer learning technology. But in the process of rapidly constructing a network without considering the specificity of the CXR image. To solve the above problem, this paper proposed a novel XDense-RC-net. Based on the DenseNet model, this method improved the capability of feature extraction and feature fusion by introducing a new spatial attention mechanism in the densely connected layer, Optimized Transition module of DenseNet by using two different pooling strategies to enhance the noise immunity of the model. XDense-RC-net used the chest X-ray14 (multilabel and 14-class) and COVIDx (single label and 3-class) datasets for validating. In the multi-label classification experiments, the average AUC score reached 0.854, which is 0.109 higher than the benchmark method. In the single-label classification experiments, the average accuracy reached 96.75%, which is 7.75% higher than the baseline model. The results show that XDense-RC-net improves the accuracy of CXR image classification and can generalize to multilabel and single-label classification tasks. [ABSTRACT FROM AUTHOR]

To solve the problem of inadequate feature extraction by the model due to factors such as occlusion and illumination in person re-identification tasks, this paper proposed a model with a joint cross-consistency learning and multi-feature fusion person re-identification. The attention mechanism and the mixed pooling module were first embedded in the residual network so that the model adaptively focuses on the more valid information in the person images. Secondly, the dataset was randomly divided into two categories according to the camera perspective, and a feature classifier was trained for the two types of datasets respectively. Then, two classifiers with specific knowledge were used to guide the model to extract features unrelated to the camera perspective for the two types of datasets so that the obtained image features were endowed with domain invariance by the model, and the differences in the perspective, attitude, background, and other related information of different images were alleviated. Then, the multi-level features were fused through the feature pyramid to concern the more critical information of the image. Finally, a combination of Cosine Softmax loss, triplet loss, and cluster center loss was proposed to train the model to address the differences of multiple losses in the optimization space. The first accuracy of the proposed model reached 95.9% and 89.7% on the datasets Market-1501 and DukeMTMC-reID, respectively. The results indicated that the proposed model has good feature extraction capability. [ABSTRACT FROM AUTHOR]

Subtle facial expression recognition (SFER) aims to classify facial expressions with very low intensity into corresponding human emotions. Subtle facial expression can be regarded as a special kind of facial expression, whose facial muscle movements are more difficult to capture. In the last decade, various methods have been developed for common facial expression recognition (FER). However, most of them failed to automatically find the most discriminative parts of facial expression and the correlation of muscle movements when human makes facial expression, which makes them unsuitable for SFER. To better solve SFER problem, an attention mechanism based model focusing on salient local regions and their correlations is proposed in this paper. The proposed method: 1) utilizes multiple attention blocks to attend to distinct discriminative regions and extract corresponding local features automatically, 2) a correlation attention module is integrated in the model to extract global correlation feature over the salient regions, and finally 3) fuses the correlation feature and local features in an efficient way for the final facial expression classification. By this way, the useful but subtle local information can be utilized in more detail, and the correlation of different local regions is also extracted. Extensive experiment on the LSEMSW and CK+ datasets shows that the method proposed in this paper achieves superior results, which demonstrates its effectiveness. [ABSTRACT FROM AUTHOR]

Rainy images typically contain heterogeneous rain distributions; however, many existing methods perform well in simple homogeneous rain and fail to handle complex heterogeneous rain effectively. In this paper, we try to solve this problem by fully exploiting the complementary contextual information in the manner of a Joint Feedback and Recurrent deraining scheme with Ensemble Learning (JFREL). First, the proposed JFREL is built on a recurrent multistage architecture, and the output of each stage is fused automatically via ensemble learning. Second, the feedback mechanism is utilized to refine information from inter- and intra-stages. Third, at each stage the residual dilated aggregation attention module is recursively adopted to adequately characterize complementary high-level contextual information in multiple receptive fields and adaptively aggregate beneficial details to achieve feature compensation. Extensive experiments demonstrate that the proposed JFREL can achieve a competitive performance over the state-of-the-art methods on both synthetic and real-world datasets. [ABSTRACT FROM AUTHOR]

Semantic segmentation of standing trees is important to obtain factors of standing trees from images automatically and effectively. Aiming at the accurate segmentation of multiple standing trees in complex backgrounds, some traditional methods have shortcomings such as low segmentation accuracy and manual intervention. To achieve accurate segmentation of standing tree images effectively, SEMD, a lightweight network segmentation model based on deep learning, is proposed in this article. DeepLabV3+ is chosen as the base framework to perform multi-scale fusion of the convolutional features of the standing trees in images, so as to reduce the loss of image edge details during the standing tree segmentation and reduce the loss of feature information. MobileNet, a lightweight network, is integrated into the backbone network to reduce the computational complexity. Furthermore, SENet, an attention mechanism, is added to obtain the feature information efficiently and suppress the generation of useless feature information. The extensive experimental results show that using the SEMD model the MIoU of the semantic segmentation of standing tree images of different varieties and categories under simple and complex backgrounds reaches 91.78% and 86.90%, respectively. The lightweight network segmentation model SEMD based on deep learning proposed in this paper can solve the problem of multiple standing trees segmentation with high accuracy. [ABSTRACT FROM AUTHOR]

In order to solve the problem of the existing research work focuses on the compatibility of paired items, this paper proposed an outfit compatibility prediction model based on hypergraph representation. The model constructs a fashion hypergraph based on the category information of fashion items and the collocation relationship between different fashion items in the existing dataset, where each hypernode represents an item, and each hyperedge represents an outfit made up of multiple items. To better infer outfit compatibility from the hypergraph, the model converts hypergraphs into traditional graphs, and the graph neural network is used to simulate the complex interaction between nodes. Finally, the attention mechanism is introduced to calculate the outfit compatibility score to strengthen the predictive ability of the model. Experimental results show that in the outfit fill-in-the-blank task and the outfit compatibility prediction task, the model achieved an accuracy rate of 77.29% and 96.23% respectively, which was significantly improved compared with other baseline models. [ABSTRACT FROM AUTHOR]

As the core technology in the field of intelligent transportation, traffic sign detection and recognition play a significant role in unmanned driving, high-precision map navigation, and auxiliary driving. However, due to the inevitable extreme weather and human impacts in the process of driving on the road, traffic sign detection and recognition remains a difficult problem. In this paper, a traffic sign detection and recognition algorithm based on the improved YOLOv4 is proposed to improve the original Feature Pyramid Network, so that the feature layer has more semantic information. Additionally, the coordinate attention mechanism module is added to the model, which fully considers the relationship between channels and position information. An adaptive feature fusion module is also added before the detection head to solve the problem of unclear semantic information caused by scaling in target detection. The experimental results indicate that the detection accuracy of the improved YOLOv4 algorithm is increased by 5%, and the detection performance is better than that of other target detection algorithms. [ABSTRACT FROM AUTHOR]

The recommender system uses knowledge graph can solve the problems of data sparsity and cold start, but it often ignores the importance of high-order collaborative information and different collaborative information to explore users' potential preferences. Therefore, this paper proposed a recommendation model combining enhanced collaborative information and knowledge graph information ( CIKG ) . This model first used the historical interactive data of users and items to obtain first-order collaboration information and high-order collaboration information. At the same time, it used the attention mechanism to capture important information and obtain enhanced collaboration information to supplement the feature representation of users and items. Secondly, this model matched interacted items with the entities in the knowledge graph and performed the propagation in the knowledge graph. It could obtain knowledge graph information and used the information to obtain the user' s preferences and enhance the interpretability of the model. Finally, the aggregator combined the enhanced collaboration information and knowledge graph information to obtain the final representation of users and items, so as to make predictions. The experimental results on Last-fm and Book-crossing datasets show that CIKG has a great improvement over other comparative models. [ABSTRACT FROM AUTHOR]

In recent years, the research of aspect level sentiment analysis attempts to model the dependency between context words and aspects by using attention mechanism and graph convolution model based on dependency tree. However, the accuracy of these methods is limited by noise information and lack of differentiated message passing mechanism. In order to solve the above problems, this paper explored a new method combining attention mechanism with syntactic knowledge, used dependency tree and location information to supervise the attention mechanism respectively, and designed and proposed a dependency tree enhanced attention model for aspect level sentiment analysis. The model could make more rational use of semantic and syntactic information and reduce the dependence on dependency tree. Experiments on three benchmark datasets verify the effectiveness and interpretability of the proposed method . [ABSTRACT FROM AUTHOR]

Pixel-level image fusion is an effective way to fully exploit the rich texture information of visible images and the salient target characteristics of infrared images. With the development of deep learning technology in recent years, the image fusion algorithm based on this method has also achieved great success. However, owing to the lack of sufficient and reliable paired data and a nonexistent ideal fusion result as supervision, it is difficult to design a precise network training mode. Moreover, the manual fusion strategy has difficulty ensuring the full use of information, which easily causes redundancy and omittance. To solve the above problems, this paper proposes a multi-stage visible and infrared image fusion network based on an attention mechanism (MSFAM). Our method stabilizes the training process through multi-stage training and enhances features by the learning attention fusion block. To improve the network effect, we further design a Semantic Constraint module and Push–Pull loss function for the fusion task. Compared with several recently used methods, the qualitative comparison intuitively shows more beautiful and natural fusion results by our model with a stronger applicability. For quantitative experiments, MSFAM achieves the best results in three of the six frequently used metrics in fusion tasks, while other methods only obtain good scores on a single metric or a few metrics. Besides, a commonly used high-level semantic task, i.e., object detection, is used to prove its greater benefits for downstream tasks compared with single-light images and fusion results by existing methods. All these experiments prove the superiority and effectiveness of our algorithm. [ABSTRACT FROM AUTHOR]

Aiming at the problem of missing detection and false detection caused by inaccurate motion feature extraction of existing video key frame extraction algorithms, this paper proposed a reinforcement learning key frame extraction algorithm combining multi-channel feature and attention mechanism. The algorithm extracted the human skeleton joint points from the video sequence through the human posture recognition algorithm firstly. Then it used the S-GCN and ResNet50 network to extract the motion features and static features in the video sequence respectively, and performed a weighted fusion of the two. Finally, it applied the attention mechanism to calculate the importance of the video frame of the feature sequence, and used reinforcement learning to extract and optimize key frames. The experimental results show that the algorithm can solve the problem of missing and false detection in the key frame extraction of motion video. It performs well in the detection of video frames containing key actions, with high accuracy and strong stability. [ABSTRACT FROM AUTHOR]

Nowadays intelligent power production uses electric power images widely. However, due to the development of image editing software, some power images are maliciously tampered, which seriously affects the power production process. In order to solve this problem, this paper proposed a dual-channel CenterNet model for power image splicing forgery detection. The color channel extracted the obvious visual features of the image. The stage-analysis channel explored the noise feature. At the same time, the algorithm used a feature fusion module based on attention mechanism, it integrated the dual-channel features to enhanced the feature recognition ability. Experiments show that the model can get better detection performance and has practical significance in the forgery detection application of power images. [ABSTRACT FROM AUTHOR]

At present, aspect-based sentiment analysis based on graph convolution neural network (GCN) integrates the grammatical structure of sentences. It can effectively solve the problem of inaccurate long-range lexical dependence. However, GCN has unnecessary complexity and redundant computing. In addition, it ignores the relative position relationship between attributes and context. This paper proposed a new model to solve these problems. Firstly, the model established a bidirectional GRU layer. Secondly, it used position-aware transformation to increase the importance of context words close to aspect words, and then reduced the complexity by removing nonlinearity and folding the weight matrix between successive layers. Thirdly, it fused the specific mask layer to realize the single-layer MASGC (masked simple graph convolutional networks) structure, and then generated a new attention mechanism based on retrieval context. Finally, it gave the classification results through the full connection layer. The model carries out a large number of experiments on five data sets. The experimental results show that it has higher accuracy and less training time. [ABSTRACT FROM AUTHOR]

Existing deep learning based single image super-resolution (SISR) models usually improve the fitting ability of the model by increasing the number of network layers, but fail to fully extract and reuse features, leading low quality of reconstructed images. To solve this problem, this paper proposed an image super-resolution model based on feature fusion and attention mechanism. This model used residual in residual (RIR) structure in feature extraction module. The feature extraction module of the network consisted of several residual groups. Each residual group consisted of several residual blocks. This module implemented local feature fusion in each residual group and global feature fusion between each group. In addition, this model introduced coordinate attention module into each residual block and spatial attention module into each residual group. It verifies that the model is able to fully extract features and reuse features. The final experimental results show that the model is superior to the existing models in objective evaluation indexes and subjective visual effect. [ABSTRACT FROM AUTHOR]

Aiming at the problem of low person re-identification accuracy caused by similar appearance and posture of person in surveillance videos, this paper proposed a video-based person re-identification method based on a graph model, which effectively utilized the time sequence information in the video to realize the information interaction across frames and intra-frame regions. Specifically, it used the correlation information between the cross-frame block regions to establish the inherent relationship between the regional nodes, and iteratively updated the regional information through feature propagation. On the other hand, in the metric learning process, it proposed a weighted loss function strategy, which optimized the binary allocation method (that is, discarding or retaining the sample) in the previous mining strategy into a continuous score allocation method, which solved the problem that the available samples were not used efficiently. Finally, it evaluated the model on MARS and DukeMTMC-VideoReID datasets, and the experimental results confirm the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]