2022 7th International Conference on Image, Vision and Computing (ICIVC)最新文献

{"title":"A Multi-scale Radar HRRP Target Recognition Method Based on Pyramid Depthwise Separable Convolution Network","authors":"Jiaxing He, Xiaodan Wang, Yafei Song, Qian Xiang","doi":"10.1109/ICIVC55077.2022.9886642","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9886642","url":null,"abstract":"Radar high resolution range profile (HRRP) contains important structural features such as target size and scattering center distribution, which has attracted extensive attention in the field of radar target recognition. In order to solve the problem of feature extraction and recognition in HRRP target recognition, we propose an HRRP target recognition method based on one-dimensional Pyramid Depthwise Separable Convolutional (PyDSC) neural network. For the processed data, pyramid convolution is selected, and convolution kernels of different sizes are used on different input channels, which can better extract the features of different scales and improve the overall recognition ability. At the same time, Depthwise Separable Convolution (DSC) technology is applied to PyConv network, a standard convolution operation is divided into two steps: deep convolution and point convolution, which can reduce the network complexity, reduce the amount of parameters and improve the speed of HRRP target recognition. Finally, we verify the effectiveness of the proposed method through experiments. The experimental results show that: 1) compared with the other three convolutional neural networks, our proposed PyDSC can significantly improve the recognition accuracy with a small increase in overhead; 2) Compared with the original PyConv, PyDSC can effectively reduce the complexity of the model.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121166170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Skeleton-Based Dumbbell Fitness Action Recognition Using Two-Stream LSTM Network","authors":"Mingzhou Shang, Qian Huang, Yiming Wang, Xiang Bian, Chuanxu Jiang, Jiwen Liu","doi":"10.1109/ICIVC55077.2022.9886880","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9886880","url":null,"abstract":"With the development of 3D skeleton extraction technology, skeleton-based action recognition has made significant progress in recent years. However, there are few studies on dumbbell fitness action recognition. Therefore, this paper collects a 3D skeleton sequence dataset based on dumbbell fitness (DUM-Action3D) and proposes an anomaly detection method based on clustering local outlier factor algorithm in the process of data sampling. In particular, in terms of feature extraction, this paper proposes a method to extract mixed multi-dimensional features for action classification and designs a hierarchical two-stream fusion LSTM network. Experiments demonstrate that our method is better than the traditional LSTM network and has a more robust capability of learning representations. Furthermore, our method achieves good recognition accuracy and execution speed on the dataset.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"181 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113968902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Anomaly Intrusion Detection in Online Learning Space Based on XGBoost","authors":"Qian Dong, Tingting Sun, Tingwei Li, Xiaoqian Lu, Kai Yan","doi":"10.1109/ICIVC55077.2022.9886479","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9886479","url":null,"abstract":"To ensure the network security of the online learning space, this paper proposes a network anomaly intrusion detection method based on XGBoost that can be applied to the online learning space. Considering about the high latitude and non-linear characteristics of abnormal network traffic data in real scenarios, this method combines multiple weak classifiers into a stronger classifier through an ensemble learning method to solve the dimensional disaster and low operation efficiency of traditional machine learning algorithms. Compared with the traditional ensemble learning methods, this method adds a regularization operator, and uses the second-order approximation of the loss function to select features at the intermediate nodes. These make the method more powerful in detection performance and efficiency. To examine the effect of the proposed method, this paper uses a public data set for validation. In the experiments, the AUC value is 0.9957, and the trend of the ROC curve points out that the method can effectively detect anomaly attacks.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"70 9","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114031835","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Research on On-line Detection Method and Device Development of HF and Micro Water Content with Optical Image Processing","authors":"Zhang Shiling","doi":"10.1109/ICIVC55077.2022.9887220","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9887220","url":null,"abstract":"In this paper, the infrared absorption spectrum characteristics of H2O and HF are studied in detail. The subsequent experimental research is carried out at the wave number of 1392nm for H2O and 1278nm for HF. A spectral analysis system based on the FTIR infrared spectrometer is established to study the spectral absorption characteristics of H2O and HF with temperature and pressure. The TDLAS experimental system was built, and the preliminary experimental research on H2O and HF gas was carried out, which reached the detection limit of ppmv level. After determining research scheme, the project team, combined with TDLAS technology, designed four kinds of optical path tanks made of aluminum alloy, stainless steel, PTFE and PVDF for gas adsorption experiments. The analysis of the experimental results shows that there are differences in adsorption saturation time of optical path tanks made of different materials, but it does not affect the accuracy of measurement results. In practical application, the adsorption of H2O and HF gas by tank materials can not be considered. With the help of TracePro ray tracing software, the Herriott multiple reflection gas absorption cell is theoretically studied and simulated. The detection experiments of H2O and HF were carried out. The results show that when using TDLAS technology combined with 1392nm laser to detect H2O, the accuracy can reach 2 ℃; When using TDLAS technology combined with 1278nm QCL laser to detect HF, the detection limit can be less than 1ppmv. The results meet the requirements of project indicators. Considering the volume, cost, the anti-interference performance, the gas consumption, convenience of assembly and commissioning and wide application of later achievements, the project team finally designed and trial produced an optical gas absorption cell in the form of pure optical fiber for on-line detection of high-voltage equipment. At present, the project team has completed the trial production of key components of the prototype, such as optical path cell unit, laser driving unit, laser temperature control unit and digital processing unit. Fill the gap of H2O and HF live detection at home and abroad, improve and improve the evaluation means of SF6 Electrical equipment operation, and further effectively ensure the safe and stable operation of the equipment.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124324471","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Object Detection Algorithm Based on Global Information Fusion","authors":"Juanjuan Li, Zhiqiang Hou, Ying Sun, Hao Guo, Sugang Ma","doi":"10.1109/ICIVC55077.2022.9886816","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9886816","url":null,"abstract":"The output channel of the Fully Convolutional One-Stage Object Detection (FCOS) feature extraction network drops sharply before input FPN. Feature information is severely lost. This paper proposes an object detection algorithm based on global information fusion. First, a multi-scale global aggregation module is designed. This module extracts the information of multi-scale receptive fields, aggregates global features, and enhances local features. The last layer features of the feature backbone network are downsampled, and fused with feature downsampling enhanced using a multi-scale global aggregation module. The enhanced feature upsampling and shallow feature fusion is output through FPN. The proposed algorithm uses Generalized Intersection over Union loss instead of Intersection over Union loss, which can make the target localization more accurate. The detection accuracy of the algorithm in this paper on the PASCAL VOC dataset reaches 82.8%, which is 2.0% higher than FCOS. The accuracy on the KITTI dataset reaches 82.4%, which is 4.2% higher than FCOS. And the detection speed meets the real-time requirements.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"142 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124542432","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ensemble Long Short-Term Tracking with ConvNeXt and Transformer","authors":"Yuhua Xiao, Yifeng Zhang, Pengyu Ni","doi":"10.1109/ICIVC55077.2022.9887117","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9887117","url":null,"abstract":"Visual object tracking is an important research topic in Computer Vision. The widely used Siamese network architecture learns a similarity metric between target objects and search regions, and locates the targets in video sequences. In this paper, we present an ensemble long short-term tracking algorithm based on ConvNeXt and Transformer. Firstly, a Siamese network with the ConvNeXt backbone is applied to extract features for both target and search regions. Secondly, an encoder-decoder transformer is introduced to capture global feature dependencies. In addition, an IoU-confidence-based tracking ensemble algorithm is designed to capture both long-term stable appearances and short-term variable appearances of the target. The proposed tracker, called STARK-NeXt, achieves a success rate of 68.9% on LaSOT, outperforming STARK by 1.8%.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121770806","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transfer Learning with Deep Convolutional Neural Network for Automated Plant Identification","authors":"Wei Liu, Huirui Han, Guilai Han","doi":"10.1109/ICIVC55077.2022.9886149","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9886149","url":null,"abstract":"Automated plant identification enables experts to process significantly greater numbers of plants with higher efficiencies in shorter periods. It is time-consuming and difficult to determine the name of species based on observations, even for botanist experts. However, plant recognition is a kind of fine-grained visual recognition problem, which is relatively harder than conventional image recognition. To solve this problem, we present a solution that transfers the learning information from a Deep Convolutional Neural Network (DCNN) trained on the ImageNet database, which contains millions of images, for automated plant identification based on flower and fruit images. First, we modify the last three layers of the pre-trained network to adapt ResNet-50 model to our classification task and replace the fully connected layer in the original pre-trained network with another fully connected layer, in which the output size represents the class of plants. Second, we use transfer experience and fine-tuned pre-trained DCNN for experiments using flower and fruit images. Finally, we evaluate the proposed network on two available botanical datasets: the Oxford flowers dataset with 102 classes and the HNPlant flowers and fruits dataset with 20 classes and determine the optimal values of the associated hyperparameters to improve the overall performance. Experiment results demonstrate that the highest classification accuracies exhibited by the proposed model on the Oxford-102 and HNPlant-20 datasets are 92.4% and 95.0%, respectively, thus establishing their effectiveness and superiority.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133657793","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Coupled Shift-Invariant Tensorial Spatial ICA Applied to Multi-Group Complex-Valued Task-Related and Resting-State fMRI Data","authors":"Li-Dan Kuang, Zhi-Ming He","doi":"10.1109/ICIVC55077.2022.9886323","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9886323","url":null,"abstract":"Multi-subject complex-valued fMRI data, inheriting high noise and spatiotemporal variability, do not well conform canonical polyadic decomposition (CPD) model. Therefore, spatial independence, phase sparsity and temporal shift-invariance have been added in CPD model, e.g., shift-invariant tensorial spatial independent component analysis (sT-sICA). However, different subjects may belong to different groups and scans, paradigms and responses may vary among groups. Considering that coupled CPD (CCPD) can preserve common and different information among different groups, we propose a novel coupled sT-sICA in the framework of shift-invariant CCPD model. We estimate shared spatial maps (SMs) by performing three-stage principal component analysis and ICA, and calculate group-specific time courses (TCs), subject-specific time delays and intensities by conducting complex-valued shift-invariant rank-one matrix approximation. Actual task-related and resting-state fMRI data experiments verify that the proposed method extracts better shared SMs and group-specific TCs and captures larger temporal differences between healthy controls and schizophrenic patients than sT-sICA and CCPD.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127052534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An Lévy Flight Based Honey Badger Algorithm for Robot Gripper Problem","authors":"J. Zhong, Xinguang Yuan, Bo Du, Gang Hu, Congyao Zhao","doi":"10.1109/ICIVC55077.2022.9887256","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9887256","url":null,"abstract":"The honey badger algorithm (HBA) is a recent meta-heuristic optimization algorithm that solves optimization problems by simulating the foraging behavior of honey badgers. To solve the poor convergence of this algorithm in the face of complex optimization problems and to improve the optimization performance of HBA, this paper proposes an enhanced Lévy based HBA algorithm and applies it to the optimization problem of the robot gripper. First, we improve the optimization efficiency of the basic HBA by using the Lévy flight strategy to enhance the local search capability and avoid falling into the local optimum. Secondly, we verify the performance of LHBA by the CEC2020 test function. The experiments show that the LHBA algorithm has good optimization ability. Finally, LHBA is used to solve the robot gripper optimization problem. The results show that LHBA can obtain the minimum value of the difference between the minimum force and the maximum force and successfully solve this optimization problem.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114356993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Study on Dielectric Property Test and Image Computer Recognition of UHVDC Bushing","authors":"Zhang Shiling","doi":"10.1109/ICIVC55077.2022.9886796","DOIUrl":"https://doi.org/10.1109/ICIVC55077.2022.9886796","url":null,"abstract":"This paper briefly describes the diagnosis methods of bushing insulation, and focuses on the research status of dielectric response measurement methods. By studying the design method of bushing, the basic insulation structure parameters of 10kV model bushing are designed, and the basic steps of how to trial produce bushing under laboratory conditions are explained. The temperature control platform and dielectric response test platform for 10kV AC epoxy resin impregnated dry bushing are built. The dielectric spectra of dry bushing and epoxy resin sample are measured respectively. Comparing the experimental results of dry bushing and epoxy resin sample, it is found that the dielectric constant and dielectric loss tangent of bushing are much larger than those of epoxy resin sample, because the epoxy resin will shrink when curing during manufacturing, And the sample sleeve used for measurement may be damp during storage; The experimental data are fitted by the model of superposition of three HN functions to verify the effectiveness of the model, and the characteristic parameters such as DC conductivity are extracted. It is found that its dependence on temperature basically conforms to Arrhenius formula. On this basis, build a thermal aging test platform for epoxy resin samples, carry out the accelerated thermal aging test at 115 °C, measure the dielectric spectrum of samples with different aging time with the dielectric response test system, compare the dielectric spectrum under different aging time, and analyze the reasons for the differences. The experimental results show that the dielectric constant will rise with the increase of aging time, The dielectric loss tangent will not change much. Combined with the image database information, insulation distance information and typical main equipment casing E-field distribution information, the operation state parameters are effectively obtained, and the intelligent algorithm is applied to automatically evaluate the operation state.","PeriodicalId":227073,"journal":{"name":"2022 7th International Conference on Image, Vision and Computing (ICIVC)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114973738","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}