2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)最新文献

{"title":"Arrhythmia classification method based on knowledge distillation and 2D ECG images","authors":"Renwei Li, L. Meng, Yian Liu, Shaogang Hu, G. Qiao","doi":"10.1109/NNICE58320.2023.10105764","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105764","url":null,"abstract":"Arrhythmia classification is crucial for identifying and preventing the potential risks associated with cardiovascular diseases. Currently., using deep learning for arrhythmia classification faces challenges of model complexity and deployment difficulty. To address this., We present a novel approach for classifying arrhythmias using knowledge distillation and 2D ECG images. In this study., the classification of the MIT-BIH dataset is performed according to AAMI criteria. Preprocessing., such as denoising., interception., and 2D grayscale image conversion., is first performed. Then., high-performance 2D-CNN models are pre-trained using 2D ECG images as the teacher model., and a lightweight 1D-CNN is constructed as the student model. Finally., we align raw and 2D ECG data., utilizing knowledge distillation to allow the student model to learn from the teacher model during the training process., thus enabling model compression. The final results show that the method significantly improves the performance of the student model in arrhythmia classification. For overall performance., when using Vgg16 as the teacher model., Acc and P+ improve by 1.62% and 6.64%., and Se is comparable. In addition., the F1-score values have been improved in all categories. The proposed method provides a feasible approach for low computational power device deployment for arrhythmia classification.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"245 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134088119","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":"Traffic Track Dynamic Data Analysis Based on BP Neural Network Model","authors":"Yue Zhang, D. Dong, Junlin Sha","doi":"10.1109/NNICE58320.2023.10105670","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105670","url":null,"abstract":"The prediction of track irregularity requires the mileage correction of the data collected by the track inspection car. Mileage correction of track data can directly affect the evaluation of railway management personnel on track diseases. Mileage correction of track data can directly affect the evaluation of railway management personnel on track diseases. In this paper, the construction of background value, the selection of initial value and the optimization model of weight proportion direction of detection data are applied to capture the trend change law of track shape with greater advantages. In the part of stochastic term development, BP neural network model is introduced to modify the residual sequence of track irregularity. The final experimental results show that the relative prediction error of the proposed model is reduced by 18. 2% and 15. 7% respectively, compared with the single gray model. The validity and robustness of the model are verified. The improved track irregularity prediction model has good prediction accuracy, and the preventive maintenance decision scheme of railway tracks can be realized according to the model.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"96 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133274379","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":"Electronic efficiency and nonlinearity of Ka-band broadband traveling wave tube","authors":"Chang-jiang Tong, Qo Bo","doi":"10.1109/NNICE58320.2023.10105687","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105687","url":null,"abstract":"The space traveling wave tube (TWT) is a vital device for satellite signal amplification, which determines the lifespan, reliability, and stability of a satellite. Space TWTs require high efficiency and linearity, and bandwidth expansion is a development prospect for TWTs. Optimizing linearity and improving electronic efficiency are crucial for broadband TWTs. This paper reports a study on the influence of the cold parameters of a Ka-band broadband TWTs on the nonlinear phase shift and electronic efficiency. The normalized phase velocity, coupling impedance, and attenuation constant of each frequency were varied to conduct the investigations. The results showed that when the dispersion and coupling curves flatten, the nonlinear phase shift between the high and low frequencies decreases, and the electronic efficiency only decreases slightly. Finally, the slow-wave structure of the T-shaped vane was used to optimize the electronic efficiency and linearity by adjusting the pitch. The electronic efficiency of the 18–24 GHz broadband TWT was greater than 18%, the nonlinear phase shift was less than 60°, and the nonlinear phase shift difference between the high and low frequencies was -12°.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131908843","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 Intrusion Detection Method Based on Transformer-LSTM Model","authors":"Zhipeng Zhang, Xiaotian Si, Linghui Li, Yali Gao, Xiaoyong Li, Jie Yuan, Guoqiang Xing","doi":"10.1109/NNICE58320.2023.10105733","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105733","url":null,"abstract":"With the development of network technologies, network intrusion has become increasing complex which makes the intrusion detection challenging. Traditional intrusion detection algorithms detect intrusion traffic through intrusion traffic characteristics or machine learning. These methods are inefficient due to the dependence of manual work. Therefore, in order to improve the efficiency and the accuracy, we propose an intrusion detection method based on deep learning. We integrate the Transformer and LSTM module with intrusion detection model to automatically detect network intrusion. The Transformer and LSTM can capture the temporal information of the traffic data which benefits to distinguish the abnormal data from normal data. We conduct experiments on the publicly available NSL-KDD dataset to evaluate the performance of our proposed model. The experimental results show that the proposed model outperforms other deep learning based models.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133276230","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":"Neural Matrix Factorization Model Based On Latent Factor Learning","authors":"Ao Chang, Qingxian Wang, Xiao Zhang","doi":"10.1109/NNICE58320.2023.10105775","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105775","url":null,"abstract":"Aiming at the typical problem of predicting users' evaluation of items in recommender systems, we propose a neural matrix factorization model based on latent factor learning. Based on the classical latent factor model, the model utilizes the representation ability of deep learning, adds auxiliary features and cross features in the process of vectorization to improve the abstract expression ability of the model to the latent factor vector, and realizes the improvement of the model by auxiliary feature information. Finally, compared with the existing deep learning models, the prediction results are significantly improved in RMSE and MAE metrics.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117174655","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":"FITrans: Skin Lesion Segmentation Based on Feature Integration and Transformer","authors":"Likun Zhang","doi":"10.1109/NNICE58320.2023.10105777","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105777","url":null,"abstract":"Automatic segmentation of skin lesions in dermatoscopic images is of great significance for clinical diagnosis and evaluation of melanoma. Most existing skin lesion segmentation methods overfit the primary visual region in skin lesion images but ignore the edge details. When the boundary of the lesion is blurred and the contrast between the lesion and the background is low, it is difficult to obtain an accurate segmentation. To solve this problem, this paper proposes a network model FITrans that combines multi-level feature aggregation and Transformer. Firstly, the convolution neural network uses additional classification information and boundary information to extract discriminative context and edge features by weakly supervised weight feedback and explicit edge guidance, respectively. Secondly, the model integrates the global-level context features, low-level edge features, and high-level semantic features extracted from the backbone network. The integrated multi-level features can compensate for the shortcomings of single-feature segmentation and enhance the segmentation effect of edge areas. On this basis, the Transformer model codes the serialized integration features, which not only make up for the problem that the Transformer does not make full use of the low-level features, but also extracts rich information on distance dependence between pixels. Combining the multi-level integration features with the Transformer can further enhance the segmentation effect of edge regions. In this paper, FITrans are used to analyze the effect of different levels of features on segmentation, and to verify the validity of multi-level feature integration and distance-dependent information. FITrans achieved 88.6 and 81.7 Jaccard coefficients (JAs) on ISIC2016 and ISIC2017, which exceed the five mainstream skin lesion segmentation methods. The experimental results show that the FITrans model can effectively improve the accuracy of target edge segmentation, and is more helpful for the accurate segmentation of complex skin damage images.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"11973 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116366562","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 The Selection And Calibration Algorithm Of Special Monitoring Camera","authors":"Yahui Zhang","doi":"10.1109/NNICE58320.2023.10105684","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105684","url":null,"abstract":"As an important medium for image processing technology to obtain original data for algorithm analysis, the camera is closely related to the accuracy of the algorithm. The first prerequisite to ensure the accuracy of the algorithm is the correct camera selection and stable installation position. Therefore, it is very necessary to establish a model that can calibrate the camera, and judge whether the installation position of the camera complies in real-time and match the camera selection guide, to make a strong guarantee for the accuracy of the algorithm.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116948229","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":"Deep Learning Based Wireless Device Identification Using RF Fingerprint","authors":"Ziang Lu, Tengyan Wang, Xin Liu, Nige Li","doi":"10.1109/NNICE58320.2023.10105692","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105692","url":null,"abstract":"The wireless network physical layer identity authentication technology has received more and more attention due to its strong resistance to camouflage attacks. This work studies a physical layer method that uses radio frequency fingerprints to distinguish legitimate devices from illegal devices. This paper proposes the constellation trajectory generated by the fusion of three different length differential intervals of short, medium, and long as the device fingerprint to solve the problem that the traditional single differential interval constellation trajectory can not take into account the frequency offset resolution and the application range of frequency offset. The experimental results show that the fingerprint recognition accuracy of devices using multiple differential interval constellation trajectories is better than traditional methods. On this basis, we propose a radio frequency fingerprint identification method based on the twin neural network. Unlike most previous studies that model the identification problem as a classification problem, the algorithm proposed in this paper detects whether the device under test is a legitimate user device and can authenticate the legal device identity. The results show that the radio frequency fingerprint identification method for wireless devices based on the twin neural network can effectively and accurately identify legal users and illegal users.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115943488","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":"A Survey of Vehicle Trajectory Prediction Based on Deep-Learning","authors":"Huilin Yin, Yurong Wen, Jiaxiang Li","doi":"10.1109/NNICE58320.2023.10105706","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105706","url":null,"abstract":"Trajectory prediction is very important for autonomous driving vehicles. Deep-learning approaches can capture the interaction information. Therefore more and more researchers are using deep-learning models to accomplish vehicle trajectory prediction. This paper provides a detailed and comparative description of the methods for vehicle trajectory prediction based on deep-learning models. We start by modelling trajectory predictions for vehicles. Several common deep-learning models used in vehicle trajectory prediction are then highlighted. Next, we compare the structural features, salient advantages and applicable scenarios of commonly used deep-learning models. Future directions for vehicle trajectory prediction are discussed at the end.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"401 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124241903","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":"Layer-wise Top-$k$ Gradient Sparsification for Distributed Deep Learning","authors":"Guang-pu Li","doi":"10.1109/NNICE58320.2023.10105718","DOIUrl":"https://doi.org/10.1109/NNICE58320.2023.10105718","url":null,"abstract":"Distributed training is widely used in training large-scale deep learning models, and data parallelism is one of the dominant approaches. Data-parallel training has additional communication overhead, which might be the bottleneck of training system. Top-$boldsymbol{k}$ sparsification is a successful technique to reduce the communication volume to break the bottleneck. However, top-$boldsymbol{k}$ sparsification cannot be executed until backpropagation is completed, which disables the overlap of backpropagation computations and gradient communications, leading to limiting the system scaling efficiency. In this paper, we propose a new distributed optimization approach named LKGS-SGD, which combines synchronous SGD (S-SGD) with a novel layer-wise top-$boldsymbol{k}$ sparsification algorithm (LKGS). The LKGS-SGD enables the overlap of computations and communications, and adapts to gradient exchange at layer-wise. Evaluations are conducted by real-world applications. Experimental results show that LKGS-SGD achieves similar convergence to dense S-SGD, while outperforming the original S-SGD and S-SGD with top-$boldsymbol{k}$ sparsification.","PeriodicalId":263114,"journal":{"name":"2023 3rd International Conference on Neural Networks, Information and Communication Engineering (NNICE)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128373887","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}