Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263544
Lijuan Duan, Yao Wang, Ying Xiao, Yuanhua Qiao, Changming Wang
Two seizure prediction models are built based on a decision fusion strategy and residual network by using spatial coupling features and introducing an attention mechanism. First, eight frequency bands are filtered, and the correlation matrices are computed for each frequency of eighteen channels. Second, the eight 18x18 matrices are input to the residual module for classification, and the results are concatenated to form a vector. A fully connected layer is used for decision fusion. Third, to emphasize the coupling relationship among the different frequency bands, a cubic matrix formed by the eight 18x18 matrices is inputted to an attention network, resulting in the enhanced features. A seizure prediction model is thus proposed by combining the nine decisions. The performance of the model is compared with those from state-of-the-art methods, and the sensitivity of the proposed model is improved by 4.45%.
{"title":"Prediction of Seizure via Residual Networks Based on Decision Fusion","authors":"Lijuan Duan, Yao Wang, Ying Xiao, Yuanhua Qiao, Changming Wang","doi":"10.1109/CISP-BMEI51763.2020.9263544","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263544","url":null,"abstract":"Two seizure prediction models are built based on a decision fusion strategy and residual network by using spatial coupling features and introducing an attention mechanism. First, eight frequency bands are filtered, and the correlation matrices are computed for each frequency of eighteen channels. Second, the eight 18x18 matrices are input to the residual module for classification, and the results are concatenated to form a vector. A fully connected layer is used for decision fusion. Third, to emphasize the coupling relationship among the different frequency bands, a cubic matrix formed by the eight 18x18 matrices is inputted to an attention network, resulting in the enhanced features. A seizure prediction model is thus proposed by combining the nine decisions. The performance of the model is compared with those from state-of-the-art methods, and the sensitivity of the proposed model is improved by 4.45%.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131017842","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263500
Yang Zhu, Zhaowei Liu, Yuanqing Ma
CP-nets are one of the powerful tools for learning uncertain relations in the field of artificial intelligence, in which learning problems are the main research content. So far, there are many learning methods about CP-nets, which are widely used in information retrieval, user decision-making, recommendation systems and other fields. This paper attempts to propose a new solution, using an algorithm based on the combination of conditional independence and scoring search to learn the structure of CP-nets, it combines local learning, constraint-based and search scoring techniques, which is principled and effective. Firstly, this paper using the MMPC algorithm to get CPC(candidate parent-child node), and then selecting the appropriate search algorithm as a measurement standard, and perform a scoring search to obtain the optimal network structure. Based on the idea of simulated annealing in the search phase, this paper combines the probability jump feature to perform random search in the solution space to avoid falling into the local optimal solution, obtain better search results, and compare with several search algorithms. In the experimental part, it is compared with other hybrid algorithms such as sparse candidate algorithm, and the agreement is used as a measurement standard to verify the effectiveness of the algorithm.
{"title":"Structure Learning of CP-nets Based on Constraint and Scoring Search","authors":"Yang Zhu, Zhaowei Liu, Yuanqing Ma","doi":"10.1109/CISP-BMEI51763.2020.9263500","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263500","url":null,"abstract":"CP-nets are one of the powerful tools for learning uncertain relations in the field of artificial intelligence, in which learning problems are the main research content. So far, there are many learning methods about CP-nets, which are widely used in information retrieval, user decision-making, recommendation systems and other fields. This paper attempts to propose a new solution, using an algorithm based on the combination of conditional independence and scoring search to learn the structure of CP-nets, it combines local learning, constraint-based and search scoring techniques, which is principled and effective. Firstly, this paper using the MMPC algorithm to get CPC(candidate parent-child node), and then selecting the appropriate search algorithm as a measurement standard, and perform a scoring search to obtain the optimal network structure. Based on the idea of simulated annealing in the search phase, this paper combines the probability jump feature to perform random search in the solution space to avoid falling into the local optimal solution, obtain better search results, and compare with several search algorithms. In the experimental part, it is compared with other hybrid algorithms such as sparse candidate algorithm, and the agreement is used as a measurement standard to verify the effectiveness of the algorithm.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131045895","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263615
Zeming Jin, Ling Wang, Hanhan Liu, Ronghua Du, Xiang Zhang
Relative pose estimation based on vision is widely used in various space navigation tasks. Considering the relative pose estimation problem of a spacecraft autonomous approaching to an unknown and non-cooperative target, a method for monocular-based pose estimation of non-cooperative space targets using Extended Kalman filter (EKF) and Extended Kalman Particle Filter (EKPF) is proposed. Compared with the existing methods, the proposed method does not depend on the prior information such as the size and shape of the target, and only uses the coordinates of the feature points of the target image as the filter input to realize the fast and accurate estimation of all pose parameters. Simulation results verify the effectiveness and feasibility of the proposed method.
{"title":"Monocular-Based Pose Estimation of Non-Cooperative Space Targets Using EKF and EKPF","authors":"Zeming Jin, Ling Wang, Hanhan Liu, Ronghua Du, Xiang Zhang","doi":"10.1109/CISP-BMEI51763.2020.9263615","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263615","url":null,"abstract":"Relative pose estimation based on vision is widely used in various space navigation tasks. Considering the relative pose estimation problem of a spacecraft autonomous approaching to an unknown and non-cooperative target, a method for monocular-based pose estimation of non-cooperative space targets using Extended Kalman filter (EKF) and Extended Kalman Particle Filter (EKPF) is proposed. Compared with the existing methods, the proposed method does not depend on the prior information such as the size and shape of the target, and only uses the coordinates of the feature points of the target image as the filter input to realize the fast and accurate estimation of all pose parameters. Simulation results verify the effectiveness and feasibility of the proposed method.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132211956","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263557
Wen-Chao Yang, Tzu-Huan Lin
Digital foreosics has developed vigorously, aod the demaod for traceability of digital imagiog equipmeo t is iocreasiog day by day. Receotly, a sigoificaot brea kthrough is achieved by usiog the Photo-Respoose Noo-Uoiformity (PRNU) ooise of images to trace the imagiog device. However, digital images are ofteo takeo with mobile phooes aod theo traosmitted usiog social media (such as LINE software io Taiwao ) io real cases. Duriog the traosmissioo process, most of the images are resized aod compressed. To explore the impact of this issue oo image traceability techoology, related experimeots are desigoed to evaluate io this study. 15 differeot Apple mobile phooes were used to iodividually capture digital images to create the data sets. After beiog traosmitted through LINE software, they were dowoloaded. The correlatioo evaluatioo method is ba sed oo the modified correlatioo eoergy peak (Modified Sigoed Peak Correlatioo Eoergy, MSPCE) statistics to evaluate aod aoalyze the correlatioo betweeo the PRNU factors of the dis puted images aod those io the origioal data sets. Experimeotal results show that the proposed method could effectively trace the source of the imagiog device usiog the distorted images which are resized aod compressed duriog the traosmissioo io LINE.
{"title":"Source camera identification in LINE social network via CCD fingerprint","authors":"Wen-Chao Yang, Tzu-Huan Lin","doi":"10.1109/CISP-BMEI51763.2020.9263557","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263557","url":null,"abstract":"Digital foreosics has developed vigorously, aod the demaod for traceability of digital imagiog equipmeo t is iocreasiog day by day. Receotly, a sigoificaot brea kthrough is achieved by usiog the Photo-Respoose Noo-Uoiformity (PRNU) ooise of images to trace the imagiog device. However, digital images are ofteo takeo with mobile phooes aod theo traosmitted usiog social media (such as LINE software io Taiwao ) io real cases. Duriog the traosmissioo process, most of the images are resized aod compressed. To explore the impact of this issue oo image traceability techoology, related experimeots are desigoed to evaluate io this study. 15 differeot Apple mobile phooes were used to iodividually capture digital images to create the data sets. After beiog traosmitted through LINE software, they were dowoloaded. The correlatioo evaluatioo method is ba sed oo the modified correlatioo eoergy peak (Modified Sigoed Peak Correlatioo Eoergy, MSPCE) statistics to evaluate aod aoalyze the correlatioo betweeo the PRNU factors of the dis puted images aod those io the origioal data sets. Experimeotal results show that the proposed method could effectively trace the source of the imagiog device usiog the distorted images which are resized aod compressed duriog the traosmissioo io LINE.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132261297","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263573
Xiaoying Wei, Yanhua Cao, Xiaozhong Yang
Image segmentation is a key technique in the image processing and a classic problem. A Monte Carlo segmentation algorithm based on the Markov Random Field (MRF) image model is proposed to randomly initialize the model parameters, so as to avoid the over-dependence of the algorithm on the initial value and overcome the shortcomings of the local optimal solution of the existing iterative algorithm. Firstly, the MRF model can make full use of the neighborhood relationship of pixel space to obtain the data field information of the image. Then according to the Bayesian theory, the prior knowledge of images is transformed into the prior distribution model. Finally, the Monte Carlo segmentation algorithm is used to iterate until the maximum posterior probability is reached, thus, the distribution of image labels is obtained, that is, the process of image segmentation is completed. The segmentation experiment shows that the Monte Carlo algorithm can overcome the shortcoming of the traditional iterative algorithm, which is trapped in the local optimal, and can segment the image in a more complete and detailed way, effectively realize the accuracy of segmentation, and improve the speed of image segmentation.
图像分割是图像处理中的一项关键技术，也是一个经典问题。提出了一种基于马尔可夫随机场(Markov Random Field, MRF)图像模型的蒙特卡罗分割算法，对模型参数进行随机初始化，避免了算法对初始值的过度依赖，克服了现有迭代算法局部最优解的缺点。首先，MRF模型可以充分利用像素空间的邻域关系获取图像的数据场信息;然后根据贝叶斯理论，将图像的先验知识转化为先验分布模型。最后，使用蒙特卡罗分割算法进行迭代，直到达到最大后验概率，从而得到图像标签的分布，即完成图像分割过程。分割实验表明，蒙特卡罗算法克服了传统迭代算法陷入局部最优的缺点，能够将图像分割得更完整、更细致，有效地实现了分割的准确性，提高了图像分割的速度。
{"title":"The Monte Carlo Algorithm for Image Segmentation Based on the MRF Model","authors":"Xiaoying Wei, Yanhua Cao, Xiaozhong Yang","doi":"10.1109/CISP-BMEI51763.2020.9263573","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263573","url":null,"abstract":"Image segmentation is a key technique in the image processing and a classic problem. A Monte Carlo segmentation algorithm based on the Markov Random Field (MRF) image model is proposed to randomly initialize the model parameters, so as to avoid the over-dependence of the algorithm on the initial value and overcome the shortcomings of the local optimal solution of the existing iterative algorithm. Firstly, the MRF model can make full use of the neighborhood relationship of pixel space to obtain the data field information of the image. Then according to the Bayesian theory, the prior knowledge of images is transformed into the prior distribution model. Finally, the Monte Carlo segmentation algorithm is used to iterate until the maximum posterior probability is reached, thus, the distribution of image labels is obtained, that is, the process of image segmentation is completed. The segmentation experiment shows that the Monte Carlo algorithm can overcome the shortcoming of the traditional iterative algorithm, which is trapped in the local optimal, and can segment the image in a more complete and detailed way, effectively realize the accuracy of segmentation, and improve the speed of image segmentation.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122684803","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263545
Shan Liu, Mingxi Li, Shicong Song, Jinge Li, Yan Yan
In the era of media convergence, the communication mode of movie and television dramas has undergone great changes from cross-screen communication to cross-platform communication. It has been a challenge to construct a comprehensive, effective, reasonable analysis of the communication effects. The purpose of this research is to construct a comprehensive cross-platform communication effect evaluation index system for movies and television dramas based on the Analytic Hierarchy Process (AHP) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). The simulation results demonstrated that our model can evaluate the comprehensive communication effect and rank movies and TV dramas accurately.
在媒介融合时代，影视剧的传播方式发生了巨大的变化，从跨屏传播到跨平台传播。如何构建全面、有效、合理的传播效果分析已成为一个挑战。本研究的目的是基于层次分析法(AHP)和TOPSIS法(Order Preference Technique by Similarity to Ideal Solution, TOPSIS)，构建一个综合性的影视剧跨平台传播效果评价指标体系。仿真结果表明，该模型能够较准确地评价影视剧的综合传播效果，并对影视剧进行排名。
{"title":"Cross-platform Communication Effect Evaluation Model for Movies and TV Dramas","authors":"Shan Liu, Mingxi Li, Shicong Song, Jinge Li, Yan Yan","doi":"10.1109/CISP-BMEI51763.2020.9263545","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263545","url":null,"abstract":"In the era of media convergence, the communication mode of movie and television dramas has undergone great changes from cross-screen communication to cross-platform communication. It has been a challenge to construct a comprehensive, effective, reasonable analysis of the communication effects. The purpose of this research is to construct a comprehensive cross-platform communication effect evaluation index system for movies and television dramas based on the Analytic Hierarchy Process (AHP) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). The simulation results demonstrated that our model can evaluate the comprehensive communication effect and rank movies and TV dramas accurately.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122927717","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263671
Jie-Cyun Huang, Xinrui Zhao, L. Zheng
Micro expression is a facial feature that can reflect the most real emotional state hidden in the human heart. This is a very short process and difficult to capture accurately. Based convolutional network, a new network architecture (SHCFNet) is proposed to extract the spatial-temporal feature of peak frames, the optical flow between onset and apex frame and its derivative (optical strain). The proposed network stacks these features from the outcomes of the previous layer. Then, the stacked feature is merged with the convolution feature of the previous layer, which enhances the learnability of neurons. The performance of the proposed SHCFNet are evaluated on four benchmark datasets: CASME I, CASME II, SAMM and SMIC, and compared with other advanced networks.
{"title":"SHCFNet on Micro-expression Recognition System","authors":"Jie-Cyun Huang, Xinrui Zhao, L. Zheng","doi":"10.1109/CISP-BMEI51763.2020.9263671","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263671","url":null,"abstract":"Micro expression is a facial feature that can reflect the most real emotional state hidden in the human heart. This is a very short process and difficult to capture accurately. Based convolutional network, a new network architecture (SHCFNet) is proposed to extract the spatial-temporal feature of peak frames, the optical flow between onset and apex frame and its derivative (optical strain). The proposed network stacks these features from the outcomes of the previous layer. Then, the stacked feature is merged with the convolution feature of the previous layer, which enhances the learnability of neurons. The performance of the proposed SHCFNet are evaluated on four benchmark datasets: CASME I, CASME II, SAMM and SMIC, and compared with other advanced networks.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127908939","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263510
Fan Meng, Xue Ni, Guang Yang, Qianqian Jia
Low-speed dim small targets are not easily detected by radar in a clutter environment. In this paper, we propose a novel approach to improve the detection probability of low-speed dim small targets, which is to convert radar data into two-dimensional images to achieve background noise suppression. Firstly, we extract the data of the target and its surroundings by setting the detection domain and make the radar data map into the data of 256 gray grades for image processing. In order to suppress clutter, we develop the improved Bilateral filter (IBF) and apply the Doppler velocity as a weight term of the Gaussian function. Combined with the weight term of spatial distance, the detection domain can be significantly enhanced. Then, the target region contour is extracted by the adaptive threshold segmentation method from the background, and the target focused is accumulated, combining with Doppler velocity. The results show that the proposed method can effectively keep the edge of the target domain and weaken the noise background, thereby improving the detection probability of the target.
{"title":"A Novel Method for Low-Speed Dim Small Target Detection","authors":"Fan Meng, Xue Ni, Guang Yang, Qianqian Jia","doi":"10.1109/CISP-BMEI51763.2020.9263510","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263510","url":null,"abstract":"Low-speed dim small targets are not easily detected by radar in a clutter environment. In this paper, we propose a novel approach to improve the detection probability of low-speed dim small targets, which is to convert radar data into two-dimensional images to achieve background noise suppression. Firstly, we extract the data of the target and its surroundings by setting the detection domain and make the radar data map into the data of 256 gray grades for image processing. In order to suppress clutter, we develop the improved Bilateral filter (IBF) and apply the Doppler velocity as a weight term of the Gaussian function. Combined with the weight term of spatial distance, the detection domain can be significantly enhanced. Then, the target region contour is extracted by the adaptive threshold segmentation method from the background, and the target focused is accumulated, combining with Doppler velocity. The results show that the proposed method can effectively keep the edge of the target domain and weaken the noise background, thereby improving the detection probability of the target.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128842371","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263601
Lecheng Weng, Xiaoqiang Li, Wenfeng Wang
In the process of a finger vein image acquisition, finger vein images are susceptible to external factors like finger posture and light source conditions, which will result in poor recognition accuracy. Therefore, a finger vein recognition method based on improved convolution neural net work is proposed to improve the accuracy and robustness of the image recognition. Firstly, the collected finger vein image is preprocessed by image segmentation, finger root key point location and image extraction in the region of interest (ROI). Secondly, according to the application context of finger vein recognition, the convolution neural network structure is adjusted appropriately, and the output of convolution layer is standardized in batches. The optimized neural network is used to automatically extract, classify and identify the features of the preprocessed images. A large number of experiments were performed on public finger print data sets of Shandong University. The optimal recognition rates are 90% respectively. The experiments verify the effectiveness of this method.
{"title":"Finger vein recognition based on Deep Convolutional Neural Networks","authors":"Lecheng Weng, Xiaoqiang Li, Wenfeng Wang","doi":"10.1109/CISP-BMEI51763.2020.9263601","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263601","url":null,"abstract":"In the process of a finger vein image acquisition, finger vein images are susceptible to external factors like finger posture and light source conditions, which will result in poor recognition accuracy. Therefore, a finger vein recognition method based on improved convolution neural net work is proposed to improve the accuracy and robustness of the image recognition. Firstly, the collected finger vein image is preprocessed by image segmentation, finger root key point location and image extraction in the region of interest (ROI). Secondly, according to the application context of finger vein recognition, the convolution neural network structure is adjusted appropriately, and the output of convolution layer is standardized in batches. The optimized neural network is used to automatically extract, classify and identify the features of the preprocessed images. A large number of experiments were performed on public finger print data sets of Shandong University. The optimal recognition rates are 90% respectively. The experiments verify the effectiveness of this method.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129111850","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}
Pub Date : 2020-10-17DOI: 10.1109/CISP-BMEI51763.2020.9263542
Liying Zhou, Xiaomin Li, Yi Liu, W. Zuo
This paper is based on literature dataset about Artificial Intelligence from SCI and EL A series of indices, such as Documents, Times Cited, CNCI, Highly Cited Papers, Hot Papers and EI Controlled Terms are used to analyze the research status and trends in the field of artificial intelligence in 2018-2020. Based on Documents, Times Cited and CNCI, high-yield countries, high-yield institutions, high-impact countries and high-impact institutions are identified. Based on Highly Cited Papers, Hot Papers and EI Controlled Terms, the most productive topics and the most influential topics in AI subject are identified. The results show that: AI is the third most productive sub-field in the Computer Science, and it produces the most highly cited papers and hot papers; the three countries with most total paper output are China mainland, USA, and Japan, while the top three countries with highest average paper impact are USA, England and United Kingdom; China mainland has the most high-yield institutions, among which Tsinghua University ranks first; the most influential topics discussed in highly cited papers are Decision Making, Neural Networks, Convolution, Fuzzy Sets, Deep Learning, Learning Algorithms, etc.
{"title":"Data Analytics for Artificial Intelligence Research from 2018 to 2020","authors":"Liying Zhou, Xiaomin Li, Yi Liu, W. Zuo","doi":"10.1109/CISP-BMEI51763.2020.9263542","DOIUrl":"https://doi.org/10.1109/CISP-BMEI51763.2020.9263542","url":null,"abstract":"This paper is based on literature dataset about Artificial Intelligence from SCI and EL A series of indices, such as Documents, Times Cited, CNCI, Highly Cited Papers, Hot Papers and EI Controlled Terms are used to analyze the research status and trends in the field of artificial intelligence in 2018-2020. Based on Documents, Times Cited and CNCI, high-yield countries, high-yield institutions, high-impact countries and high-impact institutions are identified. Based on Highly Cited Papers, Hot Papers and EI Controlled Terms, the most productive topics and the most influential topics in AI subject are identified. The results show that: AI is the third most productive sub-field in the Computer Science, and it produces the most highly cited papers and hot papers; the three countries with most total paper output are China mainland, USA, and Japan, while the top three countries with highest average paper impact are USA, England and United Kingdom; China mainland has the most high-yield institutions, among which Tsinghua University ranks first; the most influential topics discussed in highly cited papers are Decision Making, Neural Networks, Convolution, Fuzzy Sets, Deep Learning, Learning Algorithms, etc.","PeriodicalId":346757,"journal":{"name":"2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116957258","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：