Pub Date : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339713
Yi Zhang, Weizhen Zeng, Gangqiang He, Yueyuan Liu
Image captioning is a task that enables computer to naturally describe the contents of an image like a human, moreover it involves two different major research fields of computer vision and natural language processing. In this paper, a new image captioning system is proposed, which can address the challenges of automatically describing images in the wild. Built on the state-of-the-art caption framework, we designed a deep visual detector to catch a broad range of visual concepts, a GAN(Generative Adversarial Network) with graph embedding is developed to generate accurate sentences for wild images.
{"title":"SVGAN: Semi-supervised Generative Adversarial Network for Image Captioning","authors":"Yi Zhang, Weizhen Zeng, Gangqiang He, Yueyuan Liu","doi":"10.1109/TOCS50858.2020.9339713","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339713","url":null,"abstract":"Image captioning is a task that enables computer to naturally describe the contents of an image like a human, moreover it involves two different major research fields of computer vision and natural language processing. In this paper, a new image captioning system is proposed, which can address the challenges of automatically describing images in the wild. Built on the state-of-the-art caption framework, we designed a deep visual detector to catch a broad range of visual concepts, a GAN(Generative Adversarial Network) with graph embedding is developed to generate accurate sentences for wild images.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126176325","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339721
Xiaoyu Wu, Yu Wang, Naimeng Cang
In response to the unsafe driving of taxi drivers during the epidemic, the design and implementation of the driver's face registration and detection of whether the driver wears a mask, detection of the driver's fatigue physiological signals and multiple integration are designed. Target detection algorithm based on MobileNetV2 to realize mask detection. Combine MTCNN and Face-Net organically to realize driver's face login. The cerebellar neural network model is used to fuse the 12 fatigue monitoring signals EMG, EEG, etc. extracted by the simulated driving experiment platform to obtain a multi-integrated fatigue monitoring control model. After the fatigue driving multiple physiological indicators of the simulated driving platform, the technical model is studied and verified. It is concluded that the multiple fusion fatigue monitoring control model has higher accuracy than the traditional single signal monitoring.
{"title":"Detection and Research on Unsafe Driving of Taxi Drivers","authors":"Xiaoyu Wu, Yu Wang, Naimeng Cang","doi":"10.1109/TOCS50858.2020.9339721","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339721","url":null,"abstract":"In response to the unsafe driving of taxi drivers during the epidemic, the design and implementation of the driver's face registration and detection of whether the driver wears a mask, detection of the driver's fatigue physiological signals and multiple integration are designed. Target detection algorithm based on MobileNetV2 to realize mask detection. Combine MTCNN and Face-Net organically to realize driver's face login. The cerebellar neural network model is used to fuse the 12 fatigue monitoring signals EMG, EEG, etc. extracted by the simulated driving experiment platform to obtain a multi-integrated fatigue monitoring control model. After the fatigue driving multiple physiological indicators of the simulated driving platform, the technical model is studied and verified. It is concluded that the multiple fusion fatigue monitoring control model has higher accuracy than the traditional single signal monitoring.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116221010","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339700
Xingbing Ma
To gain separately controlled center frequencies and realize miniaturization, this article presents a dual-band bandpass filter (BPF) adopting narrow-closed-loop loaded resonator (NCLLR). By adjusting two narrow-side positions (namely the size of narrow-closed-loop), center frequencies of two passbands can be independently shifted in a certain area, respectively. Moreover, in order to achieve high selectivity in frequency characteristic domain, tapped-line coupling is applied to introduce three/four transmission zeroes (TZs) on either side of two target passbands. Good characteristics in passbands can be obtained by optimizing coupling gap and tapping position of stepped impedance I/O feed lines. Simulated and measured results are in good agreement.
{"title":"Dual-Band BPF with Independently Controllable Center Frequencies Using Narrow-Closed-Loop Loaded Resonator","authors":"Xingbing Ma","doi":"10.1109/TOCS50858.2020.9339700","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339700","url":null,"abstract":"To gain separately controlled center frequencies and realize miniaturization, this article presents a dual-band bandpass filter (BPF) adopting narrow-closed-loop loaded resonator (NCLLR). By adjusting two narrow-side positions (namely the size of narrow-closed-loop), center frequencies of two passbands can be independently shifted in a certain area, respectively. Moreover, in order to achieve high selectivity in frequency characteristic domain, tapped-line coupling is applied to introduce three/four transmission zeroes (TZs) on either side of two target passbands. Good characteristics in passbands can be obtained by optimizing coupling gap and tapping position of stepped impedance I/O feed lines. Simulated and measured results are in good agreement.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114912828","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339751
Minghuo Xia, Yu'an He
According to the concept and characteristics of the mass personalization production mode, combined with the concept of “Internet + Manufacturing” and related technologies of the industry 4.0 strategy, this paper proposes an overall framework of a smart factory for mass personalization production, and analyzes the key technologies and functional modules of the framework in detail. Taking the production of personalization bicycles in a bicycle intelligent manufacturing laboratory as an example to illustrate the process of achieving mass personalization production.
{"title":"Research on the Construction of Smart Factory for Mass Personalization Production","authors":"Minghuo Xia, Yu'an He","doi":"10.1109/TOCS50858.2020.9339751","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339751","url":null,"abstract":"According to the concept and characteristics of the mass personalization production mode, combined with the concept of “Internet + Manufacturing” and related technologies of the industry 4.0 strategy, this paper proposes an overall framework of a smart factory for mass personalization production, and analyzes the key technologies and functional modules of the framework in detail. Taking the production of personalization bicycles in a bicycle intelligent manufacturing laboratory as an example to illustrate the process of achieving mass personalization production.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129971022","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339705
Lu Rui
In order to solve the limitation of traditional K-means algorithm in dealing with large-scale data, a fast approximate k-means algorithm (FAKM) is proposed based on the approximate k-means algorithm (AKM) and the idea of classifying the cluster centers. The algorithm omits the cluster centers which only obtain a few samples in the AKM clustering results, and makes full use of the cluster centers with dense and stable samples in the cluster, In the iterative process, the number of samples and categories to be clustered is gradually reduced, which improves the speed of the algorithm and simplifies the clustering results. The FAKM algorithm is applied to the actual image retrieval system, and the experimental results show that the retrieval accuracy, retrieval time and clustering time of the system are greatly improved
{"title":"Application of Intelligent Clustering Algorithm in Image Processing","authors":"Lu Rui","doi":"10.1109/TOCS50858.2020.9339705","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339705","url":null,"abstract":"In order to solve the limitation of traditional K-means algorithm in dealing with large-scale data, a fast approximate k-means algorithm (FAKM) is proposed based on the approximate k-means algorithm (AKM) and the idea of classifying the cluster centers. The algorithm omits the cluster centers which only obtain a few samples in the AKM clustering results, and makes full use of the cluster centers with dense and stable samples in the cluster, In the iterative process, the number of samples and categories to be clustered is gradually reduced, which improves the speed of the algorithm and simplifies the clustering results. The FAKM algorithm is applied to the actual image retrieval system, and the experimental results show that the retrieval accuracy, retrieval time and clustering time of the system are greatly improved","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122666051","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339615
Helong Wang, Y. Jiang, Chong Qi, Junyan Li
This paper designs an integrated electromechanical servo system of control, drive and actuator to solve the technical problems of large weight and volume caused by the low degree of integration of traditional servo systems. It mainly includes the following aspects, the design composition and working principle of the integrated integrated electromechanical servo system, the simulation analysis of integrated integrated servo system, and the experimental research of integrated integrated servo system.
{"title":"Research on Integrated Electro-mechanical (EMA)Servo System","authors":"Helong Wang, Y. Jiang, Chong Qi, Junyan Li","doi":"10.1109/TOCS50858.2020.9339615","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339615","url":null,"abstract":"This paper designs an integrated electromechanical servo system of control, drive and actuator to solve the technical problems of large weight and volume caused by the low degree of integration of traditional servo systems. It mainly includes the following aspects, the design composition and working principle of the integrated integrated electromechanical servo system, the simulation analysis of integrated integrated servo system, and the experimental research of integrated integrated servo system.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123080794","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339748
Lin Liu, Yujing Xia, Lin Tang
Due to the high cost of biological data access and the privacy issues, collecting a large amount of biological data for training deep learning model is difficult in the field of biology. Concerning this issue, this article focuses on generative adversarial networks (GANs), which is a special type of deep learning model, and reviews their representative applications for generating biological data. We briefly introduced the working principle of GAN, and numerous applications to the areas of various biological data. In this paper, the types of biological data generated by GAN are categorized into two areas: biological sequences and two-dimensional data. These related studies indicated that GANs are able to explore the space of possible data configurations, and tuning the generated data to have specific target properties. This article will provide valuable insights and serve as a starting point for carrying out further studies for researchers.
{"title":"An overview of biological data generation using generative adversarial networks","authors":"Lin Liu, Yujing Xia, Lin Tang","doi":"10.1109/TOCS50858.2020.9339748","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339748","url":null,"abstract":"Due to the high cost of biological data access and the privacy issues, collecting a large amount of biological data for training deep learning model is difficult in the field of biology. Concerning this issue, this article focuses on generative adversarial networks (GANs), which is a special type of deep learning model, and reviews their representative applications for generating biological data. We briefly introduced the working principle of GAN, and numerous applications to the areas of various biological data. In this paper, the types of biological data generated by GAN are categorized into two areas: biological sequences and two-dimensional data. These related studies indicated that GANs are able to explore the space of possible data configurations, and tuning the generated data to have specific target properties. This article will provide valuable insights and serve as a starting point for carrying out further studies for researchers.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114513240","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339701
Linjun Li, Bo Wang, Xinyong Zhu, Denggang Yin
Chromatography is a technique used to separate, quantify each component in a mixture and is widely used in chemistry and biology. In experiments, the control of the segments of the equipment, the acquisition of data, and the complex analysis of data determine the results of chromatography. The chromatography workstation is a unique tool for experimenters to make the quantitative analysis. However, the existed workstations inherit the traditional frameworks and only partially introduce the database to manage the experimental process. This paper introduces a chromatography workstation that is completely built on a database and can use high-efficient data sheets to implement a thorough procedure from the control of the equipment, data acquisition to data analysis, user management, and report printing, and therefore substantially improves the efficiency of the analysis and the maintainability of data.
{"title":"A chromatography workstation built on a database and its applications","authors":"Linjun Li, Bo Wang, Xinyong Zhu, Denggang Yin","doi":"10.1109/TOCS50858.2020.9339701","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339701","url":null,"abstract":"Chromatography is a technique used to separate, quantify each component in a mixture and is widely used in chemistry and biology. In experiments, the control of the segments of the equipment, the acquisition of data, and the complex analysis of data determine the results of chromatography. The chromatography workstation is a unique tool for experimenters to make the quantitative analysis. However, the existed workstations inherit the traditional frameworks and only partially introduce the database to manage the experimental process. This paper introduces a chromatography workstation that is completely built on a database and can use high-efficient data sheets to implement a thorough procedure from the control of the equipment, data acquisition to data analysis, user management, and report printing, and therefore substantially improves the efficiency of the analysis and the maintainability of data.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"76 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132810272","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339716
S. Bose
Packet switching of data in networks is done by either the distance-vector or the link-state routing protocols. These protocols use the Bellman-Ford and the Dijkstra's algorithms respectively for the least cost path from a source base station to a destination station. For inter-network transmission, the path-vector routing protocol is in use. With progress of time, the network topologies are becoming huge in size, requiring large demand on book keeping of routing tables and transmission of the data packets dynamically to several other stations of the net-work by broadcast, increasing the load on the network. Here, assuming the router stations to be terrestrially located with links along the ground, a large network is assumed to lie on a spherical surface, and so the shortest geodesic path from source to destination becomes a great circular arc. For fast transmission, the cost of a link to a node is multicast to its neighboring nodes only for selection of the path lying as close as possible to the geodesic line between the source and the destination. As the arrival and dispatch of data packets at a nodal station occurs randomly, the cost of a link is estimated in this paper by the waiting time of a queueing process. This process at a router station is thus modeled by the Markovian M/M/c model, where c is the number of servers at the router station. If other commercial fixed charge is involved for the use of a link, then that can be included in the total cost of a link. Finally, a method of search of a mobile destination is also presented using sphericity of the network. Algorithms for the near geodesic path, costs of links as waiting times and destination search in mobile environment are clearly presented.
{"title":"Routing Algorithm in Networks on the Globe","authors":"S. Bose","doi":"10.1109/TOCS50858.2020.9339716","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339716","url":null,"abstract":"Packet switching of data in networks is done by either the distance-vector or the link-state routing protocols. These protocols use the Bellman-Ford and the Dijkstra's algorithms respectively for the least cost path from a source base station to a destination station. For inter-network transmission, the path-vector routing protocol is in use. With progress of time, the network topologies are becoming huge in size, requiring large demand on book keeping of routing tables and transmission of the data packets dynamically to several other stations of the net-work by broadcast, increasing the load on the network. Here, assuming the router stations to be terrestrially located with links along the ground, a large network is assumed to lie on a spherical surface, and so the shortest geodesic path from source to destination becomes a great circular arc. For fast transmission, the cost of a link to a node is multicast to its neighboring nodes only for selection of the path lying as close as possible to the geodesic line between the source and the destination. As the arrival and dispatch of data packets at a nodal station occurs randomly, the cost of a link is estimated in this paper by the waiting time of a queueing process. This process at a router station is thus modeled by the Markovian M/M/c model, where c is the number of servers at the router station. If other commercial fixed charge is involved for the use of a link, then that can be included in the total cost of a link. Finally, a method of search of a mobile destination is also presented using sphericity of the network. Algorithms for the near geodesic path, costs of links as waiting times and destination search in mobile environment are clearly presented.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"54 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114870257","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 : 2020-12-11DOI: 10.1109/TOCS50858.2020.9339759
Lifeng Hui
Pneumoconiosis is the most important occupational disease in China, and respiratory respirable dust is the main cause of pneumoconiosis. It can effectively reduce the incidence of pneumoconiosis by improving the monitoring and supervision level of respiratory dust concentration in the workplace. In order to solve the shortcomings of obtaining the concentration of respirable dust in mines by methods such as sampling by respirable dust samplers and numerical simulation experiments, an artificial neural network is proposed to predict the concentration of respirable dust. The factors affecting the concentration of respirable dust in coal mining face were analyzed, and the neural network structure for predicting respirable dust was established in this paper. Through training by selecting measured data, it was found that the error between the predicted result and the measured concentration was less than 15%, which was better than the error of regulations of dust measuring instruments. The results of the study have a certain reference effect on the prediction and prevention of respiratory dust in coal mines and the reduction of the incidence of pneumoconiosis.
{"title":"Prediction of Respirable Dust Concentration in Coal Mine Based on Neural Network","authors":"Lifeng Hui","doi":"10.1109/TOCS50858.2020.9339759","DOIUrl":"https://doi.org/10.1109/TOCS50858.2020.9339759","url":null,"abstract":"Pneumoconiosis is the most important occupational disease in China, and respiratory respirable dust is the main cause of pneumoconiosis. It can effectively reduce the incidence of pneumoconiosis by improving the monitoring and supervision level of respiratory dust concentration in the workplace. In order to solve the shortcomings of obtaining the concentration of respirable dust in mines by methods such as sampling by respirable dust samplers and numerical simulation experiments, an artificial neural network is proposed to predict the concentration of respirable dust. The factors affecting the concentration of respirable dust in coal mining face were analyzed, and the neural network structure for predicting respirable dust was established in this paper. Through training by selecting measured data, it was found that the error between the predicted result and the measured concentration was less than 15%, which was better than the error of regulations of dust measuring instruments. The results of the study have a certain reference effect on the prediction and prevention of respiratory dust in coal mines and the reduction of the incidence of pneumoconiosis.","PeriodicalId":373862,"journal":{"name":"2020 IEEE Conference on Telecommunications, Optics and Computer Science (TOCS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125282057","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}