Most Down Articles

Published in last 1 year| In last 2 years| In last 3 years| All| Most Downloaded in Recent Month | Most Downloaded in Recent Year|

Published in last 1 year

Please wait a minute...


For Selected: Download Citations EndNote Ris BibTeX Toggle Thumbnails

Select

A Moving Object Detection Algorithm Aiming at Jittery Drone Videos LIU Yaoxin1, CHEN Renxi2, YANG Weihong1 Computer and Modernization    2024, 0 (05): 99-103.   DOI: 10.3969/j.issn.1006-2475.2024.05.017 Abstract （136）      PDF（pc） （2681KB）（281）       Knowledge map    Save Abstract： To solve the problem that moving object detection is susceptible to jitter in hovering drones， leading to the generation of a significant amount of background noise and lower accuracy， a multiscale EA-KDE （MEA-KDE） background difference algorithm is proposed. This algorithm initially achieves a multiscale decomposition of image sequences to obtain a multiscale image sequence. Subsequently， before performing detection， the segmentation threshold for detection is calculated and updated by considering the area threshold and the current image frame， thereby incorporating information from the current frame. Background difference operations using high and low dual segmentation thresholds are performed on images at different scales to enhance detection robustness. Finally， a top-down fusion strategy is employed to merge the detection results from various scales， preserving the clear contours of the targets while eliminating noise. Furthermore， a proposed boundary expansion fusion post-processing algorithm helps alleviate the fragmented targets caused by detection breaks. Experimental results demonstrate that the proposed algorithm effectively suppresses background noise caused by jitter. On two real drone datasets， average F1 scores of 0.951 and 0.952 were obtained， representing improvements of 0.144 and 0.276， respectively， compared to the original algorithm. Reference | Related Articles | Metrics | Comments（0）

Select

3D Visualization Monitoring System of Aluminum Reduction Cell Based on Digital Twin ZHANG Gaoyi1, XU Yang1, 2, CAO Bin1, 3, LI Yifei3 Computer and Modernization    2024, 0 (05): 104-109.   DOI: 10.3969/j.issn.1006-2475.2024.05.018 Abstract （93）      PDF（pc） （3159KB）（271）       Knowledge map    Save Abstract：The traditional management of aluminum electrolytic cell has some problems， such as single management mode， low transparency and weak form of parameter data presentation. In order to solve these problems， digital twin technology is introduced and applied to aluminum electrolytic cell. Based on the theoretical model and framework of digital twin， a six-dimensional model of three-dimensional visual monitoring system of digital twin aluminum electrolytic cell is proposed. Based on this model， the virtual model， scene optimization， data acquisition and data mapping of electrolytic cell are constructed. The data interface is provided by Java background， and the model and data are rendered by using three.js three-dimensional technology and JavaScript language. The system provides more intuitive display effect for field personnel to better understand the operation of aluminum electrolytic cell， and provides effective ideas for the intelligent development of aluminum industry. Reference | Related Articles | Metrics | Comments（0）

Select

An Improved YOLOv5-based Method for Dense Pedestrian Detection Under Complex Road Conditions SUN Ruiqi1, DOU Xiuchao2, LI Zhihua1, JIANG Xuemei2, SUN Yuhao1 Computer and Modernization    2024, 0 (05): 85-91.   DOI: 10.3969/j.issn.1006-2475.2024.05.015 Abstract （116）      PDF（pc） （2884KB）（242）       Knowledge map    Save Abstract： Aiming at the problem of low pedestrian detection accuracy in complex street scene environment， a new network YOLO-BEN is proposed based on the improvement of YOLOv5 network. The network uses a residual connection module Res2Net with hierarchical system to integrate with C3 module，enhancing fine-grained multi-scale feature representation. The paper adopts the Bi-level routing attention module to construct and prune a region level directed graph， and applies fine-grained attention in the union of routing regions， enabling the network to have dynamic query aware sparsity and improving the feature extraction ability of fuzzy images. We incorporate the EVC module to preserve local corner area information and compensate for the problem of information loss caused by occluded pedestrians. In this paper， NWD metric and original IoU metric are used to form a joint loss function， and a small target detection head is added to improve the effect of long-distance pedestrian detection. In the experiment， the method has achieved good results on self-made data sets and some WiderPerson data sets. Compared with the original network， the accuracy， recall and average accuracy of the improved network are increased by 2.8， 4.3 and 3.9 percentage points respectively. Reference | Related Articles | Metrics | Comments（0）

Select

Review of Fall Detection Technologies for Elderly WANG Mengxi, LI Jun Computer and Modernization    2024, 0 (08): 30-36.   DOI: 10.3969/j.issn.1006-2475.2024.08.006 Abstract （400）      PDF（pc） （2530KB）（231）       Knowledge map    Save  With the rapidly growing aging population in China， the proportion of the elderly living alone has significantly increased， and thus the aging-population-oriented facilities have received increased attention. In a domestic environment， the elderly are likely to fall down due to different reasons such as lack of care， aging， and sudden illness， which have become one of the main threats to their health. Therefore， monitoring， detecting and predicting fall down behavior of the elderly in real-time can ensure their safety to some extent， while further reducing the life and health risks caused by accidental falling down. Based on a comprehensive overview of the research on human fall detection， we categorize fall detection into two categories： vision-free technologies and computer vision based methods， depending on different kinds of sensors used for data acquisition. We summarize and introduce the system composition of different methods and explore the latest relevant research， and discuss their method characteristics and practical applications. In particular， we focus on reviewing the deep learning based schemes which have been developing rapidly in recent years， while analyzing and discussing relevant principles and research results of deep learning based schemes in details. Next， we also introduce public benchmarking datasets for human fall detection， including dataset size and storage format. Finally， we discuss the prospect for the relevant research， and come up with reasonable suggestions in different aspects. Reference | Related Articles | Metrics | Comments（0）

Select

Survey on Gesture Recognition and Interaction WEI Jiakun, WANG Jiarun Computer and Modernization    2024, 0 (08): 67-76.   DOI: 10.3969/j.issn.1006-2475.2024.08.012 Abstract （300）      PDF（pc） （1322KB）（207）       Knowledge map    Save Gesture recognition and interaction technology is the cornerstone task of frontier research in human-computer interaction technology and artificial intelligence technology. This task takes the collaborative work of computers and devices to recognize and process gesture information and give machine operations corresponding to gestures as the main goal， and integrates a number of technologies such as motion capture， image processing， image classification， and multi-terminal collaborative interaction， which is a powerful guarantee to support the command and control system， robot interaction， medical operation and other cutting-edge intelligent interaction and human-computer interaction work nowadays. At present， the research on gesture recognition and interaction has become more and more mature with a wide range of application fields and rich application scenarios. This paper mainly provides a review of the gesture recognition development and interaction related technologies and hardware. Firstly， it sorts the research progress of gesture recognition and interaction technology out comprehensively， and categories the key steps of gesture recognition at the same time. Secondly， it classifies and elaborates the related work of the current mainstream gesture recognition depth sensors used for 3D gesture interaction. Subsequently， it analyses and discusses the real sense recognition technology for 3D gesture recognition. Finally， it analyses the deficiencies and urgent problems in gesture recognition and interaction technology， proposes the integration of such cutting-edge technologies as deep learning， pattern recognition and other feasible research ideas and methods， and makes predictions and prospects for the future research direction， technology development and application areas in this field. Reference | Related Articles | Metrics | Comments（0）

Select

Stance Detection with LoRA-based Fine-tuning General Language Model HAN Xiaolong, ZENG Xi, LIU Kun, SHANG Yu Computer and Modernization    2025, 0 (01): 1-6.   DOI: 10.3969/j.issn.1006-2475.2025.01.001 Abstract （245）      PDF（pc） （2429KB）（196）       Knowledge map    Save  Stance detection is a key task in natural language processing， which determines the stance of an author based on text analysis. Text stance detection methods transition from early machine learning methods to BERT models， and then evolve to the latest large language models such as ChatGPT. Distinguishing from the closed-source feature of ChatGPT， this paper proposes a text stance detection model， ChatGLM3-LoRA-Stance， by using the domestic open-source ChatGLM3 model. In order to apply large models in professional vertical fields， this paper uses LoRA efficient fine-tuning method. Compared with P-Tuning V2 efficient fine-tuning method， LoRA is more suitable for zero-shot and few-shot text stance detection tasks in text. The paper uses the publicly available VAST dataset to fine-tune the ChatGLM3 model， evaluating the performance of existing models in zero-shot and few-shot scenarios. Experimental results indicate that ChatGLM3-LoRA-Stance model has significantly higher F1 scores than other models on zero-shot and few-shot detection tasks. Therefore， the results verify the potential of large language models on text stance detection tasks， and suggest that that the use of LoRA efficient fine-tuning technology can significantly improve the performance of ChatGLM3 large language model in text stance detection tasks. Reference | Related Articles | Metrics | Comments（0）

Select

Multi-layer Bank-enterprise Converged Network Based on Graph Neural Network LI Shan, WANG Linna, GAO Dingjia, XUAN Haibo Computer and Modernization    2024, 0 (05): 27-32.   DOI: 10.3969/j.issn.1006-2475.2024.05.006 Abstract （104）      PDF（pc） （1346KB）（194）       Knowledge map    Save Abstract： The potential systemic risk in the financial industry is difficult to be accurately identified. Based on the loan data of the direct systemic risk contagion channel and internet text information of the indirect channel， a multi-layer bank-enterprise network is constructed， and a multi-layer bank-enterprise network convergence model is designed by using graph convolutional neural networks （GCN）. Based on the converged network， this paper quantitatively evaluates the systemic risk contagion process of 29 banks and 75 real estate institutions. The converged network analysis shows that the systemic risk transmission capacity under the joint impact of multi-layer bank-enterprise network is significantly greater than the systemic risk of single or two-layer network， and the systemic risk of the inter-enterprise network based on the indirect channel is more obvious. Financial prudential supervision should pay more attention to the ability of data analysis， deep learning and other technologies to integrate big data financial resources and effectively improve the level of risk monitoring and warning. Reference | Related Articles | Metrics | Comments（0）

Select

Cost-sensitive Convolutional Neural Network for Encrypted Traffic Classification#br# #br# ZHONG Hailong1, 2, HE Yueshun1, HE Linlin1, CHEN JIE1, TIAN Ming3, ZHENG Ruiyin4 Computer and Modernization    2024, 0 (05): 55-60.   DOI: 10.3969/j.issn.1006-2475.2024.05.010 Abstract （87）      PDF（pc） （1046KB）（174）       Knowledge map    Save Abstract: This paper addresses classification bias and low recognition rates for minority classes in encrypted traffic classification arising from imbalanced data. Traditional convolutional neural networks tend to favor the majority class in such scenarios， prompting a dynamic weight adjustment strategy. In this approach， during each training iteration， sample weights are adaptively adjusted based on feedback from the cost-sensitive layer. If a minority class sample is misclassified， its weight increases， urging the model to focus on such samples in future training. This strategy continually refines the model’s predictions， enhancing minority class recognition and effectively tackling class imbalance. To prevent overfitting， an early stopping strategy is employed， halting training when validation performance deteriorates consecutively. Experiments reveal that the proposed model significantly excels in addressing class imbalance in encrypted traffic classification， achieving accuracy and F1 scores over 0.97. This study presents a potential solution for encrypted traffic classification amidst class imbalance， contributing valuable insights to network security. Reference | Related Articles | Metrics | Comments（0）

Select

An Attention Mechanism-based U-Net Fundus Image Segmentation Algorithm ZHANG Zixu, LI Jiaying, LUAN Pengpeng, PENG Yuanyuan Computer and Modernization    2024, 0 (05): 110-114.   DOI: 10.3969/j.issn.1006-2475.2024.05.019 Abstract （83）      PDF（pc） （3307KB）（171）       Knowledge map    Save Abstract： The radius and width of retinal fundus vessels are important indicators for assessing eye diseases， so accurate segmentation of fundus images is becoming increasingly meaningful. In order to effectively assist doctors in diagnosing eye diseases， the paper proposes a new neural network to segment fundus vascular images. The basic idea is to reduce the information loss by improving the traditional U-Net model with the help of an attention fusion mechanism， using Transformer to construct a channel attention mechanism and a spatial attention mechanism， and fusing the information obtained by the two attention mechanisms. In addition， the number of retinal fundus images is relatively small， and the coefficients of the neural network are relatively large， which are prone to overfitting during training， so the DropBlock layer is introduced to solve this problem. The proposed method was validated on the publicly available dataset DRIVE and compared with several state-of-the-art methods. The results show that our method achieved the highest ACC value of 0.967 and the highest F1 value of 0.787. These experimental results demonstrate that the proposed method is effective in segmenting retinal fundus images. Reference | Related Articles | Metrics | Comments（0）

Select

Transmission Line Faults Detection Algorithm Based on YOLOX WU Hengfeng, HOU Xingsong, WANG Huake Computer and Modernization    2024, 0 (05): 5-10.   DOI: 10.3969/j.issn.1006-2475.2024.05.002 Abstract （131）      PDF（pc） （2405KB）（168）       Knowledge map    Save Abstract：Power system is an important foundation of national life， intelligent detection of transmission line faults has great social and economic value. Aiming at the problem of lack of public datasets in transmission line faults detection scenarios， poor performance when there are multiple scale targets simultaneously， and difficulty in detecting high IoU bounding boxes， a transmission line faults detection method based on improved YOLOX was proposed. First， a transmission line faults detection dataset was set up through acquisition and simulation; then an adaptive multi-scale feature fusion method was proposed to fully use multi-scale features; finally a new loss was proposed to improve the optimization ability of the network for high IoU bounding boxes and solve sample imbalance problem， which effectively improved the detection accuracy. The experimental results show that in the dataset collected in this paper， the proposed algorithm can still achieve 67.48% mAP50:95 while ensuring real-time performance， outperforming the classical algorithms such as EfficientDet and YOLOV5. Reference | Related Articles | Metrics | Comments（0）

Select

An LLM-based Method for Automatic Construction of Equipment Failure Knowledge Graphs ZHANG Kun1, ZHANG Yongwei1, WU Yongcheng1, ZHANG Xiaowen2, ZHAI Shichen2 Computer and Modernization    2024, 0 (11): 46-53.   DOI: 10.3969/j.issn.1006-2475.2024.11.008 Abstract （222）      PDF（pc） （5470KB）（163）       Knowledge map    Save Fault operation and maintenance is an important research topic in the field of industrial production. The research of fault prediction， fault diagnosis， question-answering systems based on the fault knowledge graph have been greatly developed and applied. However， a high-quality fault operation and maintenance knowledge graph is the foundation for these methods. Considering that traditional knowledge graph construction techniques require data preprocessing， entity recognition， relationship extraction and entity alignment of raw data， to improve the efficiency of knowledge graphs， this paper focuses on using large language models for unsupervised knowledge extraction from fault operation and maintenance data to achieve automatic construction of large-scale fault operation and maintenance knowledge graphs. This method mainly includes two parts： 1） Two zero-shot Prompts oriented towards the construction of fault operation and maintenance knowledge graphs are proposed. These Prompts can guide large language models to generate conceptual layers and extract elemental knowledge for the fault operation and maintenance knowledge graph represented and output in RDF syntax; 2） A method based on large language models for constructing knowledge graphs is proposed. This method can guide large language models to extract knowledge from fault operation and maintenance data through zero-shot Prompts and complete the construction of large-scale fault operation and maintenance knowledge graphs iteratively. Experimental results show that the proposed method is scientific and effective. Reference | Related Articles | Metrics | Comments（0）

Select

Placenta Ultrasound Image Segmentation XU Cheng1, ZHANG Yun2, ZENG Xiangjin1 Computer and Modernization    2024, 0 (05): 115-119.   DOI: 10.3969/j.issn.1006-2475.2024.05.020 Abstract （68）      PDF（pc） （2234KB）（152）       Knowledge map    Save Abstract： The shape and size of the placenta in early pregnancy are closely related to clinical outcomes such as fetal growth. Aiming at the time-consuming interactive segmentation method for three-dimensional ultrasound （3DUS） detection of placental size， a new deep learning segmentation network， DEC-U-Net， is designed based on the U-Net architecture. In the U-Net downsampling stage， deep hyperparametric convolution is used instead of 2D convolution and combined with the ECA attention mechanism. However， the accuracy of placenta detailed feature recognition is improved while introducing more parameter quantities. The cross attention mechanism is introduced into jump linking to solve the problems of blurred placental boundaries and uneven contrast. Compared with ordinary U-Net networks， the algorithm in this paper improves the intersection and merge ratio （IoU）， recall rate （Recall）， accuracy （Precision）， and Dice coefficient by 4.14， 9.59， 6.2， and 16.41 percentage points， respectively. The experimental results show that the improved network model has a good segmentation effect and can accurately segment the placenta in ultrasound images. Reference | Related Articles | Metrics | Comments（0）

Select

Underwater Trash Detection Method Based on Improved YOLOv5 PANG Mei, WANG Gong, ZHAN Yong, HUANG Zhefa Computer and Modernization    2024, 0 (07): 120-126.   DOI: 10.3969/j.issn.1006-2475.2024.07.018 Abstract （165）      PDF（pc） （3845KB）（147）       Knowledge map    Save To address the limitations of underwater image acquisition such as insufficient light， high noise and unclear object recognition， which lead to the ineffectiveness of existing object detection algorithms， an underwater garbage object detection algorithm based on improved YOLOv5 is proposed. The purpose of the improved object detection algorithm is to achieve more accurate detection and removal of underwater plastic trash from the ocean. The improved algorithm containes some improvements：using the Contrast Limited Adaptive Histogram Equalization（CLAHE） algorithm to enhance data features， which reduces the difficulty of feature extraction and enables the network to be detected more flexibly and more accurately； introducing a parameter-free attention module SimAM， using the lightweight convolution method GSConv to enhance network extraction capability while reducing model computation; At the same time， multi-scale feature fusion detection is added to solve the problem of small target location of underwater debris. Numbers of experiments are conducted based on MarineTrash which is a self-built real underwater environmental litter dataset， the results show that the improved method has good performance， in which the accuracy is increased by 4.3 percentage points， the mAP is increased by 3.5 percentage points， the GFLOPs is reduced by 0.3， and the model weight is only 13.9 MB， which is 0.6 MB lower than the baseline. The research on the underwater trash detection algorithm based on the improved YOLOv5 provides sufficient technology for deploying and installing detectors in Autonomous Underwater Vehicles （AUVs） to achieve detection and automatic removal of marine underwater trash and maintain the marine ecosystem. Reference | Related Articles | Metrics | Comments（0）

Select

Recommendation Algorithm Model Based on DNN and Attention Mechanism ZHOU Chao, CONG Xin, ZI Lingling, XIAO Guping Computer and Modernization    2024, 0 (06): 1-7.   DOI: 10.3969/j.issn.1006-2475.2024.06.001 Abstract （136）      PDF（pc） （916KB）（146）       Knowledge map    Save Abstract： In order to solve the defect of factorization machine in extracting high-order combination features and learn more useful feature information better， this paper attempts to use factorization machine to extract cross-feature and learn key feature information from low and high-order combination features by combining attention network， deep neural network， multi-head self-attention mechanism and other methods. Finally， the weighted fusion results were obtained according to the importance of the combination features of different orders， and the click-through rate of advertisements was estimated. The experiment was mainly carried out based on the advertising data set Criteo， and the analogy experiment was carried out with MovieLens data set to verify the effectiveness of the proposed algorithm model. The experimental results showed that compared with the benchmark model， in the two data sets， the AUC index increased by 2.32 percntage points and 0.4 percntage points. Reference | Related Articles | Metrics | Comments（0）

Select

Intent-based Lightweight Self-Attention Network for Sequential Recommendation HE Sida, CHEN Pinghua Computer and Modernization    2024, 0 (12): 1-9.   DOI: 10.3969/j.issn.1006-2475.2024.12.001 Abstract （125）      PDF（pc） （578KB）（143）       Knowledge map    Save  The parameters of the self-attention calculation mechanism in the existing sequence recommendation models are too large， and there is insufficient preference information in the user's shopping intention. This paper proposes an intent-based lightweight self-attention network for sequential recommendation. On the basis of the traditional product sequence embedding， the model introduces intention sequence embedding to further explore the conversion patterns between sequences. At the same time， in order to reduce the computational complexity of self-attention between pairwise products/intentions in the sequence， a convolutional segmentation sampling module is designed to divide the user behavior sequence and intention sequence into multiple segments， mapping user interests to multiple sequence segments. Comparative experiments are conducted on three public datasets， MovieLens-1M， Yelp， and Amazon-Books. Compared with baseline models， the self-attention mechanism is applied to capture the dependency between segments， effectively reducing the number of parameters. The results show that the hit rate， normalized discounted cumulative gain and mean reciprocal ranking are increased by 5.32%， 4.40% and 5.51% on the MovieLens-1M dataset， 30.93%， 22.73% and 28.84% on the Yelp dataset， and 7.78%， 11.55% and 13.98% on the Amazon-Books dataset， which verify the effectiveness of the model proposed in this paper. Reference | Related Articles | Metrics | Comments（0）

Select

Enhanced Big Language Model Dual Carbon Domain Services Based on Knowledge Graph QI Jun1, 2, QU Ruiting2, JIAO Chuanming2, ZHOU Qiaoni2, GUO Yanliang3, TAN Wenjun3 Computer and Modernization    2024, 0 (09): 8-14.   DOI: 10.3969/j.issn.1006-2475.2024.09.002 Abstract （167）      PDF（pc） （1796KB）（139）       Knowledge map    Save With the continuous development of the large language model， it has been widely applied in many fields. Due to the lack of knowledge in the dual carbon field in the big language model， the accuracy of the response results is low if the large language model is directly applied to the field of double carbon. Therefore， the method of constructing dual carbon knowledge graph as a knowledge base is adopted to enhance the application of large language models in the field of carbon peaking and carbon neutrality. The LoRA method is used to fine-tune the large language model to improve its ability to extract keywords in the carbon peaking and carbon neutrality fields. A dual carbon knowledge graph is constructed as local knowledge base to provide dual carbon domain knowledge for the model. The knowledge is used as the context of the problem， allowing the large language model to learn， and a prompt engineering assistance model is designed to generate responses. Finally， the effectiveness of the responses is evaluated. The experimental results show that， compared with the direct use of large language model， the method based on knowledge graph to enhance the dual carbon domain service of large language model has a high accuracy of intelligent response results in the field of carbon peaking and carbon neutrality， and provides an effective assistance for the construction of carbon peaking and carbon neutrality. Reference | Related Articles | Metrics | Comments（0）

Select

Knowledge Concept Recommendation Based on Meta-path and Attentional Feature Fusion LIU Yumeng, ZI Lingling, CONG Xin Computer and Modernization    2024, 0 (05): 38-45.   DOI: 10.3969/j.issn.1006-2475.2024.05.008 Abstract （96）      PDF（pc） （1350KB）（139）       Knowledge map    Save Abstract： In the research of course recommendation， the most of research effort was focused on course or video resource recommendation， only few studies paid attention to the interest or need of users for specific knowledge concept. Existing researches focus primarily on homogeneous graphs， are vulnerable to the problems of user-item relationships sparsity. To copy with the sparsity problem and fully utilize the characteristics of MOOCs datasets with multiple entities and a lot of semantic information in context relationships， a knowledge concept recommendation algorithm based on meta-path and attentional feature fusion was proposed. First， we extracted the content features of each entity and the context features between entities， input the adjacency matrices based on selected meta-paths into the graph convolutional network， and learned the representation of users and concepts under the guidance of the attention mechanism of the two-layer network structure that integrated the feature vectors of the meta-path and potential feature vectors of users and concepts. Finally， these learned user and concept representations were incorporated into an extended matrix factorization framework to predict the preference of concepts for each user. Experimental results on MOOCCube dataset demonstrate that the algorithm attains the best hit rate， the best normalized discounted cumulative gain and the best mean reciprocal ranking than those of BPR， FISM， NAIS， Metapath2vec， and MOOCIR algorithms. The algorithm improves the interpretability and prediction accuracy of the recommendation process to a certain extent， and alleviates the problem of user-item relationships sparisty. Reference | Related Articles | Metrics | Comments（0）

Select

Survey on Multimodal Information Processing and Fusion Based on Modal Categories HUANG Wendong, WANG Yifan Computer and Modernization    2024, 0 (07): 47-62.   DOI: 10.3969/j.issn.1006-2475.2024.07.008 Abstract （153）      PDF（pc） （1939KB）（139）       Knowledge map    Save  With the continuous advancement of artificial intelligence and deep learning technologies， research in the field of multimodal information processing and fusion has garnered widespread attention from researchers. This paper provides a comprehensive overview of the development history and milestone works of multimodal information processing， along with strategies and models for multimodal fusion. Based on different modalities，mainstream datasets for multimodal information processing and fusion are systematically classified and summarized. Using modality type as the classification criterion， this paper systematically reviews the research progress in multimodal information processing and fusion， emphasizing the distinctions between different modalities. Multimodal information processing and fusion are categorized into four types： audio-visual processing and fusion， audio-text processing and fusion， visual-text processing and fusion， and visual-audio-text processing and fusion. Detailed investigations are conducted on methods and models for processing and fusing different input modalities. Finally， a summary and outlook on the development of multimodal processing and fusion are provided. Reference | Related Articles | Metrics | Comments（0）

Select

Layout Analysis Method of Multi-scale Feature Fusion QIAO Jia, XU Kun, HU Peirong Computer and Modernization    2024, 0 (05): 16-21.   DOI: 10.3969/j.issn.1006-2475.2024.05.004 Abstract （105）      PDF（pc） （7439KB）（137）       Knowledge map    Save Abstract： Aiming at the problems of list and text misclassification， the difficulty of recognizing small-scale text in tables， and the poor preservation of spatial features in the current document layout element analysis， according to bottom-up thinking， the paper proposes a multi-feature fusion layout analysis method based on SegNet network. In this paper， the MSCAN-SE module is introduced into SegNet to solve the problem of low recognition rate of small-scale elements in tables. The strip features in the attention mechanism MSCAN-SE are used to improve the extraction ability of multi-scale features of the model， so that the network can retain feature information of more scales. Aiming at the problem that the features of list elements and text elements are too similar， the receptive field of the network in the feature extraction process is expanded through the dilated convolution and channel attention branch in the attention mechanism MSCAN-SE. The performance of the proposed method is compared with the classical semantic segmentation network through experiments. The results show that the pixel accuracy of the proposed method on the test set of layout analysis is 97.9%， and the mean intersection over union ratio is 91.7%. Compared with U-Net semantic segmentation model， FCN semantic segmentation model， DeepLabV3+ semantic segmentation model， and SegNet semantic segmentation model， the mean intersection and union ratio is increased by 7.6%， 2.4%， 2.6%and 1.5% respectively. Reference | Related Articles | Metrics | Comments（0）

Select

OFDM Channel Estimation Based on Matrix Recovery ZHANG Jingjing, HUANG Xuejun Computer and Modernization    2024, 0 (05): 1-4.   DOI: 10.3969/j.issn.1006-2475.2024.05.001 Abstract （117）      PDF（pc） （1380KB）（131）       Knowledge map    Save Abstract： Orthogonal frequency division multiplexing （OFDM） is a crucial technology in channel estimation， this paper proposes an OFDM channel estimation method based on matrix recovery， multiple consecutive OFDM signal in the frequency domain channel is constructed to a channel matrix. Since this channel matrix is low rank， the channel estimation problem can be converted to the weighted truncated kernel norm minimization problem of the channel matrix and the improved Singular Value Thresholding algorithm is used for recovery. The simulation results show that compared with the traditional channel estimation algorithm， the proposed method can use fewer pilot signals when the same precision channel estimation is obtained. Compared with the channel estimation method based on compressed sensing， the proposed method consumes the same amount of pilot frequency but can directly obtain high precision frequency domain estimation of OFDM channel. Reference | Related Articles | Metrics | Comments（0）

Select

Improved Roadside Monocular View Small Target Detection Algorithm Based on YOLOv5 WEI Xuecheng1, JIANG Lingyun1, LI Yan2, HE Fei2 Computer and Modernization    2024, 0 (10): 27-34.   DOI: 10.3969/j.issn.1006-2475.2024.10.005 Abstract （100）      PDF（pc） （2402KB）（131）       Knowledge map    Save Aiming at the problems of low recognition accuracy and fewer features of long-distance targets and small targets in the roadside view under vehicle-road cooperative sensing traffic scenarios， an improved algorithm for small target detection based on YOLOv5 is proposed. Firstly， in the backbone network， the GAM attention module is added to enhance the feature extraction ability of the network. Secondly， RepBi-PAN is introduced to replace the PANet structure of the original neck network to increase the network’s ability to localize small targets. Finally， the use of SIoU loss function instead of the original CIoU loss function can effectively avoid the arbitrary matching of the prediction frames in the regression process， thus enhancing the robustness of the model and accelerating the training speed of the network model. The experimental results show that compared with the original YOLOv5 6.0 version， the average accuracy mAP of each category is improved by 6.9 percentage points when the intersection over union IoU is 0.5， and the average accuracy mAP of each category is improved by 6.4 percentage points when the intersection over union IoU is 0.95， which effectively improves the detection capability of small target detection in the road-side view. Reference | Related Articles | Metrics | Comments（0）

Select

Collaborative Recommendation Algorithm with Implicit Roles YU Tianyi, LI Jianfeng, CHEN Hailong, ZHAI Jun Computer and Modernization    2024, 0 (09): 1-7.   DOI: 10.3969/j.issn.1006-2475.2024.09.001 Abstract （132）      PDF（pc） （1594KB）（130）       Knowledge map    Save This article aims to improve the effectiveness of the algorithm， starts from the psychological needs of users， locates the implicit role group of users， and researches the personalized recommendation algorithms. From a theoretical point of view， the research in this paper effectively ensures the diversity requirements of recommendation systems and improves the accuracy of algorithms to a certain extent. It expands the relevant theory of implicit preference to address the phenomenon of preference evolution. Through verification in real data， multiple experimental evaluation indicators have been significantly improved. This not only provides a theoretical basis and reference for recommendation systems， but also improves the accuracy of recommendation results. It has broad application prospects. From a practical point of view， the classification of users in this article is no longer limited to ordinary social attributes， but can further explore users’ psychological needs， obtains more accurate and diverse recommendation results， improves user satisfaction and experience. Enterprises can guide users to change their interests， increase their loyalty and value， improve their lifecycle， and increase their profits. Reference | Related Articles | Metrics | Comments（0）

Select

Automated Essay Scoring Method Based on GCN and Fine Tuned BERT MA Yu, YANG Yong, REN Ge, Palidan Tuerxun Computer and Modernization    2024, 0 (09): 33-37.   DOI: 10.3969/j.issn.1006-2475.2024.09.006 Abstract （121）      PDF（pc） （1152KB）（127）       Knowledge map    Save  Automatic scoring of essays is one of the important research directions in the field of smart education. It has the advantages of improving scoring efficiency， reducing labor costs， and ensuring the objectivity and consistency of scoring， so it has broad application prospects in the field of education. Although syntactic features play an important role in automatic scoring of compositions， there is still a lack of research on how to better utilize these features for automatic scoring of compositions. This paper proposes an automatic essay scoring method GFTB based on GCN and fine-tuned BERT. This model uses graph convolutional network to extract syntactic features of compositions， uses BERT and Adapter training methods to extract deep semantic features of compositions， and uses a gating mechanism to further capture the semantic features after the fusion of the two. The experimental results show that the proposed GFTB model achieves good average performance on 8 subsets of the public data set ASAP. Compared with baseline models such as Tongyi Qianwen， the proposed method can effectively improve the performance of automatic essay scoring. Reference | Related Articles | Metrics | Comments（0）

Select

Improved Deciduous Tree Nest Detection Method Based on YOLOv5s CHENG Meng, LI Hao Computer and Modernization    2024, 0 (08): 24-29.   DOI: 10.3969/j.issn.1006-2475.2024.08.005 Abstract （122）      PDF（pc） （2245KB）（127）       Knowledge map    Save To address the difficulty of detecting small bird nest targets in complex backgrounds， an improved YOLOv5s network architecture named YOLOv5s-nest is proposed. YOLOv5s-nest incorporates several enhancements: a refined attention mechanism called Bi-CBAM is inserted into the Backbone to effectively enhance the network’s perception of small targets; the SDI structure is introduced into the Neck to integrate more hierarchical feature maps and higher-level semantic information; the InceptionNeXt structure is inserted into the Neck to improve the model's performance and computational efficiency; and in the detection head， ordinary convolutions are replaced with PConv to efficiently extract spatial features and enhance detection efficiency. The experimental results show that the average precision of the improved model reached 89.1%， representing an increase of 6.8 percentage points compared to the original model. Reference | Related Articles | Metrics | Comments（0）

Select

Survey on Group-level Emotion Recognition in Images GAO Shuaipeng, WANG Yifan Computer and Modernization    2024, 0 (08): 98-107.   DOI: 10.3969/j.issn.1006-2475.2024.08.016 Abstract （200）      PDF（pc） （1434KB）（127）       Knowledge map    Save  In recent years， image-based group emotion recognition has received widespread attention， which aims to accurately determine the overall emotional state of groups in different scenes and with different numbers of people. Since group emotion recognition involves the analysis and fusion of multiple group emotion clues such as facial emotional features， scene features， and human posture features in pictures， this field is very challenging. At this stage， there is a lack of relevant review articles in this field to sort out the existing research， so as to better conduct the next step of research. This article carefully sorts out and categorizes group emotion recognition models with different emotional cues and different processing methods in this field. At the same time， the processing methods and characteristics of existing models are reviewed and analyzed， and models with different fusion methods and mainstream databases in this field are sorted out. Finally， a brief summary and outlook on the development of this field are given. Reference | Related Articles | Metrics | Comments（0）

Select

Multiple Unmanned Aerial Vehicles Three-dimensional Cooperative Route Planning Based on Improved GWO Algorithm JIAO Jian, JI Yuanfa, SUN Xiyan, WU Jianhui, LIANG Weibin Computer and Modernization    2024, 0 (10): 1-6.   DOI: 10.3969/j.issn.1006-2475.2024.10.001 Abstract （131）      PDF（pc） （2694KB）（127）       Knowledge map    Save To overcome the problems of poor cooperation， immersing local minimization， low convergence speed and poor solving accuracy in solving the collaborative route by GWO algorithm for multiple unmanned aerial vehicles， an improved GWO-based three-dimensional collaborative route planning algorithm for multiple unmanned aerial vehicles is proposed. Firstly， a three-dimensional collaborative trajectory planning mathematical model for multiple unmanned aerial vehicles is established， using the weighted sum of consumption cost， height cost， threat cost， spatial constraint， time constraint， and penalty term as the objective function. Secondly， the Greedy algorithm and Tent mapping are combined to improve the fitness of the population and preserve the diversity of the population to reduce the possibility of falling into local optima； then we optimize the convergence factor to improve the rate of convergence of the algorithm. Afterwards， we design a dynamic weight position update method to enhance the exploration and development capabilities of the algorithm. Finally， the improved GWO algorithm is applied to solve the trajectory planning problem of multiple unmanned aerial vehicles， and compared with GWO algorithm and CSGWO algorithm. The simulation results indicate that the proposed improved GWO algorithm enhance the solution accuracy by 64.8% and 16.7%， as well as the convergence speed by 28.5% and 25.4%， respectively. Additionally the synergy ability is significantly better than that of the comparison algorithms. Reference | Related Articles | Metrics | Comments（0）

Select

Multi-view Reconstruction with Local Self-attention and Deep Optimization YE Senhui, WANG Lei Computer and Modernization    2024, 0 (05): 92-98.   DOI: 10.3969/j.issn.1006-2475.2024.05.016 Abstract （71）      PDF（pc） （3447KB）（127）       Knowledge map    Save Abstract: To address the issues of high memory and time consumption， low completeness and fidelity of high-resolution reconstruction in multi-view 3D reconstruction， we propose a deep learning-based multi-view reconstruction network. The network consists of a feature extraction module， a cascaded Patchmatch module and a depth map optimization module. First， we design a U-shaped feature extraction module to extract multi-stage feature maps， and introduce local self-attention layers with relative position encoding at each stage， which capture the local details and global context in the images， and enhance the feature extraction performance of the network. Second， we design a deep residual network to fuse the features， and fully utilize the color image prior knowledge to constrain the depth map， and improve the accuracy of depth estimation. We test our network on the public dataset DTU （Technical University of Denmark）， and the experimental results show that our network achieves significant improvement in 3D reconstruction quality. Compared with PatchmatchNet， our network improves the completeness by 6.1% and the overall by 2.5%. Compared with other SOTA （State-Of-The-Art） methods， our network also achieves better performance in both completeness and overall. Reference | Related Articles | Metrics | Comments（0）

Select

Network Intrusion Detection Based on Improved XGBoost Model SU Kaixuan Computer and Modernization    2024, 0 (06): 109-114.   DOI: 10.3969/j.issn.1006-2475.2024.06.018 Abstract （126）      PDF（pc） （472KB）（125）       Knowledge map    Save Abstract： In order to enhance the accuracy and practicability of the traditional network intrusion detection model， this paper proposes a network intrusion detection based on an improved gradient lift tree （XGBoost） model. Firstly， the random forest algorithm is used to predict the key feature points， and the feature with the highest importance weight is effectively selected and the feature set is constructed in the data pre-processing stage. Secondly， the prediction method of XGBoost model is improved by using card equation. Finally， the cost sensitive function is introduced into the XGBoost optimization algorithm to improve the detection rate of small sample data， and the mesh method is used to reduce the complexity of the model. Experimental simulation results show that compared with other artificial intelligence algorithms， the proposed model can reduce the waiting time by more than 50% with higher inspection accuracy， and has strong scalability and adaptability under noisy environment. Combined with other models， the experimental results show that the tree depth has the greatest impact on the model performance. Reference | Related Articles | Metrics | Comments（0）

Select

Recognition of Hypopigmented Skin Diseases Based on Improved MobileNetV3-Small GAO Geng1, XIAO Fengli2, YANG Fei1 Computer and Modernization    2024, 0 (05): 120-126.   DOI: 10.3969/j.issn.1006-2475.2024.05.021 Abstract （84）      PDF（pc） （1576KB）（120）       Knowledge map    Save Abstract： In traditional hypopigmented skin disease diagnosis， reliance on the subjective clinical experience of dermatologists makes it challenging to ensure timely and accurate diagnoses for every patient. Therefore， there is a pressing need for a rapid， experience-independent diagnostic approach. Convolutional neural network （CNN） exhibits robust feature recognition capabilities， offering a potential solution. Currently， CNN -based diagnostic methods mainly focus on deeper models such as ResNet50. While achieving high accuracy， these models suffer from drawbacks like large parameter sizes， slow inference， and limited usability on mobile devices. To address these issues， this study introduces a novel lightweight CNN model based on MobileNetV3-Small. Firstly， it eliminates the computationally complex Squeeze-and-Excitation （SE） modules found in MobileNetV3-Small， replacing them with more lightweight Efficient Channel Attention （ECA） attention mechanism. Secondly， it employs the convenient and stable Leaky-ReLU activation function. Lastly， it introduces dilated convolutions in the convolutional layers to expand the receptive field. Experimental results indicate that the proposed model significantly reduces parameter size， recognition time and FLOPs compared to existing diagnostic models. It meets the high usability demands of mobile applications while still outperforming in terms of accuracy and F1 score. Ultimately， based on the proposed model， a mobile application for clinical diagnosis of hypopigmented skin disease has been developed. Reference | Related Articles | Metrics | Comments（0）

Select

Chinese Paper Invoice Text Recognition Method with Character Blurring LAI Kun Computer and Modernization    2024, 0 (08): 114-119.   DOI: 10.3969/j.issn.1006-2475.2024.08.018 Abstract （100）      PDF（pc） （1686KB）（118）       Knowledge map    Save  This paper addresses the problem of low OCR recognition performance caused by character blurring in paper invoices. A novel adaptive iterative visual semantic model is proposed to tackle this issue. The model consists of two modules： the recognition module utilizes ResNet as the encoder and Transformer as the decoder to make initial predictions on the blurred text. The correction module takes the recognition module’s predictions and feeds them into a bidirectional language model， which leverages contextual semantic information to refine characters， thereby performing initial text correction. The results are then input to a discriminator， which outputs them directly if successful or iterates the language model for further refinement if failed， effectively improving the recognition accuracy. Experimental results demonstrate that the proposed model outperforms the current state-of-the-art Chinese recognition model ch_PP-OCRv3 by 3.39 percentage points in recognition accuracy and achieves an average 6.81 percentage points improvement compared to other models. Moreover， the model exhibits excellent generalization performance on public datasets such as IC15， IIIT5K， and IC03-Word， validating its effectiveness. Reference | Related Articles | Metrics | Comments（0）

Select

Wind Power Prediction Method Based on STAGCN-Informer Spatiotemporal Fusion Model YANG Shaozu1, 2, WANG Haicheng1, 2, WU Jinya1, 2, MA Jiying1, 2 Computer and Modernization    2024, 0 (07): 13-20.   DOI: 10.3969/j.issn.1006-2475.2024.07.003 Abstract （143）      PDF（pc） （3491KB）（115）       Knowledge map    Save  Aiming at the problem that the spatial information cannot be effectively extracted due to the influence of spatiotemporal fluctuation and randomness in wind power forecasting， resulting in insufficient prediction accuracy， a model named STAGCN-Informer-DCP is proposed based on Variational Mode Decomposition （VMD），fusion of Spatiotemporal Attention Graph Convolutional Network （STAGCN） and improved Informer combination model. Firstly， VMD is used to perform modal decomposition on the original features， and the feature information on different time scales is extracted. At the same time， the selection of core parameters （penalty factor and K value） of VMD is optimized by using northern goshawk optimization （NGO）. Secondly， the STAGCN module that integrates spatio-temporal attention is used to dynamically capture the spatio-temporal features of the target wind turbine and its neighbors， and fuses them with the original signal components to obtain a feature vector carrying spatial scale information. Finally， the improved Informer model is used to extract the long-term dependencies of temporal context and realizes multi-step output prediction. The experimental results show that the combination model can better capture the dynamic space-time dependence， and effectively improve the accuracy of medium and long-term wind power forecasting. Reference | Related Articles | Metrics | Comments（0）

Select

Circular Convolutional Neural Network-based Defect Detection Method for#br# Drainage Pipe Networks LIU Cunli1, LEI Zhanzhan2, ZHENG Ao2 Computer and Modernization    2024, 0 (07): 26-35.   DOI: 10.3969/j.issn.1006-2475.2024.07.005 Abstract （111）      PDF（pc） （4477KB）（115）       Knowledge map    Save Municipal drainage systems are critical to the safety of urban road traffic， so it is important to assess their condition. In developed countries， closed-circuit television （CCTV） is the main detection tool for sewer assessment and maintenance， but it brings new challenges for its data processing. This paper proposes a drainage network defect detection method based on recurrent convolutional neural network （RCNN）. The RCNN uses a residual network （ResNet） as feature extraction module to extract visual features of drainage network image sequences， and a bidirectional LSTM is used to learn to identify temporal features to accomplish drainage network defect classification task. The method recognizes image sequences as a whole， and the training set， validation set and test set contain a total of 8800 image sequences， and 211200 images. The data set are trained and tested by the RCNN model， and the highest accuracy rate of the test set is 90.3%. Six sets of control experiments are carried out with four different fusion methods introduced to the proposed method， the SVM-based method and the method based on single frames， as well as three fusion methods based on visual attention mechanism are introduced into the proposed method and control tests are carried out. The experimental results show that the highest accuracy （90.3%） of the fusion experiments is achieved by RCNN taking the average value， and the feasibility analysis of engineering applications is realized， and the recall rate of RCNN reaches 0.977， which confirms the feasibility of the proposed method in engineering applications. Reference | Related Articles | Metrics | Comments（0）

Select

Academic Recommendation System Based on Knowledge Graph and Semantic Information ZHANG Yue, LI Huayu, ZHANG Zhikang, SHEN Xinyi Computer and Modernization    2025, 0 (01): 50-58.   DOI: 10.3969/j.issn.1006-2475.2025.01.009 Abstract （129）      PDF（pc） （1930KB）（115）       Knowledge map    Save In the diverse domains of the Internet， facing the ever-increasing volume of data， there is a growing need for recommendation systems to provide users with personalized information. Utilizing knowledge graphs can enhance the accuracy， diversity， and interpretability of these systems. Addressing the current methods’ limitations in accurately capturing genuine user preferences during propagation， and their lack of attention to the utility of semantic information， this paper proposes an advanced paper recommendation algorithm based on semantic features and knowledge graphs. This method employs the BERT model to extract semantic features from paper abstracts， and uses knowledge graphs for collaborative propagation to obtain entity representations of users and items. During propagation， user preferences are accurately transmitted through multi-head attention， and an attention aggregation network is differentiated between entity representation sets at each layer， the importance of initial information is emphasized. Performance evaluations on three public datasets demonstrate that the model proposed in this paper， compared to the selected optimal baseline models， achieves an increase of 0.010、0.018  and 0.007 in AUC， and 0.007 、0.008 and 0.008 in F1 score， respectively， thereby showing the effectiveness and the superiority of the algorithm proposed in this paper.  Reference | Related Articles | Metrics | Comments（0）

Select

Construction of Depression Recognition Model Based on Multi-Feature Fusion HOU Menghan, WEI Changfa Computer and Modernization    2025, 0 (03): 1-5.   DOI: 10.3969/j.issn.1006-2475.2025.03.001 Abstract （103）      PDF（pc） （1144KB）（114）       Knowledge map    Save  In recent years， depression has become the primary problem of global mental health burden. In order to identify it， this paper proposes a depression recognition model combining BERT， BiLSTM and ConvNeXt. Firstly， the BERT model is used to generate feature vectors with rich semantics. Secondly， the BiLSTM， and ConvNeXt model is used to obtain the context information and the local features of the text， respectively. Thirdly， to alleviate the loss of semantic information in the feature extraction process， the context and local learned by BiLSTM and ConvNeXt models are fused through residual connections. Finally， depression is recognized according to the fused feature information. The experimental results show that the proposed model improves the accuracy， recall and F1 value compared with other deep learning models， which  can effectively extract the depression features of the text and improve the accuracy of depression recognition. Reference | Related Articles | Metrics | Comments（0）

Select

Parking Positioning Method for Automatic Guided Vehicle Based on MA-LM Algorithm ZHANG Yuanchao1, 2, 3, YANG Guizhi1, 2, XUE Guang1, 3, YAO Hanchen3, PENG Jianwei3, DAI Houde3 Computer and Modernization    2024, 0 (05): 11-15.   DOI: 10.3969/j.issn.1006-2475.2024.05.003 Abstract （103）      PDF（pc） （1624KB）（114）       Knowledge map    Save Abstract： To address the challenge that autonomous navigation parking and charging solutions have poor positioning accuracy at long distances， resulting in AGVs not being able to align with the charging pile in automatic charging back mode， a parking positioning method based on an improved mayfly optimization algorithm （MA-LM） is proposed. This method fuses the magnetic nail positioning data from multiple magnetic sensor arrays， thereby improving the position accuracy and attitude accuracy of the parking positioning. To quantitatively evaluate the improvement effect of magnetic nail localization， this method is tested in a charging pile scenario using a sensor array of nine magnetic sensors and a two-wheeled differential speed mobile robot. Compared with the genetic optimization algorithm （GA-LM） and the particle swarm optimization algorithm （PSO-LM）， the experimental results show that the MA-LM algorithm has the localization accuracy of ±1.65 mm and the orientation accuracy of 0.9° in the parking localization. Reference | Related Articles | Metrics | Comments（0）

Select

Automated Drawing Psychoanalysis Based on Image Classification ZHAO Xiaoming, PAN Ting, LIU Weifeng Computer and Modernization    2024, 0 (08): 92-97.   DOI: 10.3969/j.issn.1006-2475.2024.08.015 Abstract （151）      PDF（pc） （3358KB）（114）       Knowledge map    Save  Drawing psychoanalysis method is widely used in the discovery and treatment of psychological illness and mental disorders. The House-Tree-Person （HTP） test is the most representative drawing psychoanalysis method， which projects the individual’s psychological state through the houses， trees， and persons drawn by the patient. Compared with the psychological health questionnaire， it has the advantages of being non-verbal， projective， and creative， and can systematically release the subconscious. At present， the HTP test is tested and evaluated by the therapist， which takes a long time in large-scale psychological screening， and the evaluation results will be affected by the experience and subjectivity of the therapist. Therefore， it is necessary to establish an automated method to improve the objectivity， reliability， and efficiency of the HTP test. The paper proposes an automated drawing screening method for the HTP test based on the relationship between psychological states and drawing features. The method extracts key features such as the size， position， and shadow of the drawing， and combines these features to build a psychological state classifier. This method can effectively screen out negative drawings for further diagnosis and treatment. At the same time， this paper collects the test drawing of HTP from the psychological counseling centers of the college and makes HTP dataset for experiments. Experimental results prove the superiority and application value of this method. Reference | Related Articles | Metrics | Comments（0）

Select

Research and Application of Key Technologies in Meteorological Service Middle Platform FENG Xian1, 2, FANG Kun1, QU Youming1, LIU Xiaobo1, SHI Jiachi1, WEN Liheng1 Computer and Modernization    2024, 0 (05): 69-74.   DOI: 10.3969/j.issn.1006-2475.2024.05.012 Abstract （111）      PDF（pc） （1376KB）（112）       Knowledge map    Save Abstract： With the continuous growth of meteorological data and the expansion of application scenarios， traditional data processing models are difficult to meet the needs of various industry and integrated services. In order to solve the difficulties of mass data， complex processing， demand diversification， and high response time requirements in meteorological services， we developed the Hunan meteorological service middle platform based on distributed architecture， and introduced key technologies to support high concurrency services， including adopting standardized processes to achieve unified processing of multi-source heterogeneous data， developing microservice parallel processing modules to improve data processing efficiency， designing dynamic load balancing algorithms to enhance concurrency capabilities， and ensuring operational stability through flow control mechanisms. The test results show that with the application of the above technology and limited basic resource support， the platform can support 5000 concurrent access， displaying average response time 1202 ms. It has achieved positive application effects in supporting cross industry and multi scenario meteorological services such as emergency management， water conservancy， natural resources. Reference | Related Articles | Metrics | Comments（0）

Select

Knowledge Prompt Fine-tuning for Event Extraction LI Lu, ZHU Yan Computer and Modernization    2024, 0 (07): 36-40.   DOI: 10.3969/j.issn.1006-2475.2024.07.006 Abstract （119）      PDF（pc） （1020KB）（112）       Knowledge map    Save  Event extraction is an important research focus in information extraction， which aims to extract event structured information from text by identifying and classifying event triggers and arguments. Traditional methods rely on complex downstream networks， require sufficient training data， and perform poorly in situations where data is scarce. Existing research has achieved certain results in event extraction using prompt learning， but it relies on manually constructed prompts and only relies on the existing knowledge of pre-trained language models， lacking event specific knowledge. Therefore， a knowledge based fine-tuning event extraction method is proposed. This method adopts a conditional generation approach， injecting event information to provide argument relationship constraints based on existing pre-trained language model knowledge， and optimizing prompts using a fine-tuning strategy. Numerous experiment results show that compared to traditional baseline methods， this method outperforms the baseline method in terms of trigger word extraction and achieves the best results in small samples. Reference | Related Articles | Metrics | Comments（0）

Select

Retinal Vessel Segmentation Based on Improved U-Net with Multi-feature Fusion FU Lingli, QIU Yu, ZHANG Xinchen Computer and Modernization    2024, 0 (06): 76-82.   DOI: 10.3969/j.issn.1006-2475.2024.06.013 Abstract （119）      PDF（pc） （1564KB）（112）       Knowledge map    Save Abstract： Due to some problems such as uneven distribution of blood vessel structure， inconsistent thickness， and poor contrast of blood vessel boundary， the image segmentation effect is not good， which cannot meet the needs of practical clinical assistance. To address the problem of breakage of small vessels and poor segmentation of low-contrast vessels， a CA module was integrated into the down-sampling process based on U-Net. Additiondly， to solve the problem of insufficient feature fusion in the original model， Res2NetBlock module was introduced into the model. Finally， a cascade void convolution module is added at the bottom of the model to enhance the receptive field， thereby improving the network’s spatial scale information and the contextual feature perception ability. So the segmentation task achieves better performance. Experiments on DRIVE， CHASEDB1 and self-made Dataset100 datasets show that the accuracy rates are 96.90%， 97.83% and 94.24%， respectively. The AUC is 98.84%， 98.98%， and 97.41%. Compared with U-Net and other mainstream methods， the sensitivity and accuracy are improved， indicating that the vessel segmentation method in this paper has the ability to capture complex features and has higher superiority. Reference | Related Articles | Metrics | Comments（0）

Select

Automatic Scoring Method for Composition Based on Semantic Feature Fusion YUAN Hang, YANG Yong, REN Ge, Palidan Turson Computer and Modernization    2024, 0 (06): 8-13.   DOI: 10.3969/j.issn.1006-2475.2024.06.002 Abstract （113）      PDF（pc） （623KB）（110）       Knowledge map    Save Abstract: Automatic composition scoring technology is a kind of natural language processing technology using machine learning. At present， end-to-end models based on deep learning have been widely used in the field of automatic essay scoring. However， because of the difficulty in obtaining correlations between different features in end-to-end models， Automatic Scoring Method for Composition Based on Semantic Feature Fusion （TSEF） has been proposed. This method is mainly divided into two stages: feature extraction and feature fusion. In the feature extraction stage， the Bert model is used to pre-train the input text， and a multi-head-attention mechanism is used to self-train the input text to supplement the shortcomings of pre-training; In the feature fusion stage， cross fusion methods are used to fuse the different features obtained in order to obtain a better performance model. In the experiment， TSEF was compared with many strong baselines， and the results demonstrated the effectiveness and robustness of our method. Reference | Related Articles | Metrics | Comments（0）