Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152618
Yanjun Dong, Haoyuan Hu, Min Zhu, Pan Hu, Lihong Jiang, Hongming Cai
The three-dimensional (3D) curved plate manufacturing is performed by constructing surfaces corresponding to the shape of the curved plate for multi-point forming. However, in the manufacturing process, the rebound restricts the forming accuracy, and the currently adopted rebound control methods cannot predict the rebound amount accurately. Meanwhile, the process involves multi-role collaboration and multiple data conversions and comparisons. These problems lead to a high degree of manual dependence, which affects manufacturing efficiency and accuracy. To address the above problems, this paper proposes a collaborative platform for the intelligent manufacturing of curved plates based on graph matching. Firstly, this paper establishes information models covering the whole process of curved plate manufacturing and forms a unified topology graph model. Then, the intelligent generation method of processing parameters based on graph matching is proposed, which realizes similar case recommendation and case-based processing parameters generation. Finally, we design and develop a collaboration platform based on micro-service architecture to support efficient collaboration among various departments and roles. In this paper, we use sail-shaped curved plates as a case of processing parameters generation and verify that this intelligent method can improve the accuracy of rebound control by comparison with related work, which shows that our method can be effectively applied to curved plate manufacturing.
{"title":"Intelligent Manufacturing Collaboration Platform for 3D Curved Plates Based on Graph Matching","authors":"Yanjun Dong, Haoyuan Hu, Min Zhu, Pan Hu, Lihong Jiang, Hongming Cai","doi":"10.1109/CSCWD57460.2023.10152618","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152618","url":null,"abstract":"The three-dimensional (3D) curved plate manufacturing is performed by constructing surfaces corresponding to the shape of the curved plate for multi-point forming. However, in the manufacturing process, the rebound restricts the forming accuracy, and the currently adopted rebound control methods cannot predict the rebound amount accurately. Meanwhile, the process involves multi-role collaboration and multiple data conversions and comparisons. These problems lead to a high degree of manual dependence, which affects manufacturing efficiency and accuracy. To address the above problems, this paper proposes a collaborative platform for the intelligent manufacturing of curved plates based on graph matching. Firstly, this paper establishes information models covering the whole process of curved plate manufacturing and forms a unified topology graph model. Then, the intelligent generation method of processing parameters based on graph matching is proposed, which realizes similar case recommendation and case-based processing parameters generation. Finally, we design and develop a collaboration platform based on micro-service architecture to support efficient collaboration among various departments and roles. In this paper, we use sail-shaped curved plates as a case of processing parameters generation and verify that this intelligent method can improve the accuracy of rebound control by comparison with related work, which shows that our method can be effectively applied to curved plate manufacturing.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"52 4 1","pages":"1650-1655"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83551151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152825
Han Yue, Rucen Wang, Yi Gao, Ailing Xia, Jianhua Zhang
Internet of Things (IoT) based industrial defect detection has attracted more and more attention. As a key component of intelligent manufacturing, defect detection is very important. Although deep learning (DL) can reduce the cost of traditional manual inspection and improve accuracy and efficiency, it requires huge computing resources and cannot be simply deployed on IoT devices. Digital signal processor (DSP) is an important IoT device with the characteristics of small size, strong performance and low energy consumption, and has been widely used in intelligent manufacturing. In order to achieve accurate defect detection on DSP, we proposed a variety of optimization strategies, and then extended the model to run on multi-core using a parallel scheme, and further quantified the implementation of the model. We evaluated it on three datasets, i.e. NEUSDD, MTDD and RSDD. Experimental results show that our method achieves a faster speed than running the same CNN model on a mainstream desktop CPU, with slightly accuracy loss.
{"title":"DSP-Based Industrial Defect Detection for Intelligent Manufacturing","authors":"Han Yue, Rucen Wang, Yi Gao, Ailing Xia, Jianhua Zhang","doi":"10.1109/CSCWD57460.2023.10152825","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152825","url":null,"abstract":"Internet of Things (IoT) based industrial defect detection has attracted more and more attention. As a key component of intelligent manufacturing, defect detection is very important. Although deep learning (DL) can reduce the cost of traditional manual inspection and improve accuracy and efficiency, it requires huge computing resources and cannot be simply deployed on IoT devices. Digital signal processor (DSP) is an important IoT device with the characteristics of small size, strong performance and low energy consumption, and has been widely used in intelligent manufacturing. In order to achieve accurate defect detection on DSP, we proposed a variety of optimization strategies, and then extended the model to run on multi-core using a parallel scheme, and further quantified the implementation of the model. We evaluated it on three datasets, i.e. NEUSDD, MTDD and RSDD. Experimental results show that our method achieves a faster speed than running the same CNN model on a mainstream desktop CPU, with slightly accuracy loss.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"58 1","pages":"1667-1672"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84210388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152758
Chaoyi Yan, Xinli Huang, H. Gu, Siyuan Meng
Knowledge graph embedding technique is one of the mainstream methods to handle the link prediction task, which learns embedding representations for each entity and relation to predict missing links in knowledge graphs. In general, previous convolution-based models apply convolution filters on the reshaped input feature maps to extract expressive features. However, existing convolution-based models cannot extract the interaction information of entities and relations among the same and different dimensional entries simultaneously. To overcome this problem, we propose a novel convolution-based model (SimulE), which utilizes two paths simultaneously to capture the rich interaction information of entities and relations. One path uses 1D convolution filters on 2D reshaped input maps, which maintains the translation properties of the triplets and has the ability to extract interaction information of entities and relations among the same dimensional entries. Another path employs 3D convolution filters on the 3D reshaped input maps, which is suitable for capturing the interaction information of entities and relations among the different dimensional entries. Experimental results show that SimulE can effectively model complex relation types and achieve state-of-the-art performance in almost all metrics on three benchmark datasets. In particular, compared with baseline ConvE, SimulE outperforms it in MRR by 2.9%, 9.8% and 2.8% on FB15k-237, YAGO3-10 and DB100K respectively.
{"title":"SimulE: A novel convolution-based model for knowledge graph embedding","authors":"Chaoyi Yan, Xinli Huang, H. Gu, Siyuan Meng","doi":"10.1109/CSCWD57460.2023.10152758","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152758","url":null,"abstract":"Knowledge graph embedding technique is one of the mainstream methods to handle the link prediction task, which learns embedding representations for each entity and relation to predict missing links in knowledge graphs. In general, previous convolution-based models apply convolution filters on the reshaped input feature maps to extract expressive features. However, existing convolution-based models cannot extract the interaction information of entities and relations among the same and different dimensional entries simultaneously. To overcome this problem, we propose a novel convolution-based model (SimulE), which utilizes two paths simultaneously to capture the rich interaction information of entities and relations. One path uses 1D convolution filters on 2D reshaped input maps, which maintains the translation properties of the triplets and has the ability to extract interaction information of entities and relations among the same dimensional entries. Another path employs 3D convolution filters on the 3D reshaped input maps, which is suitable for capturing the interaction information of entities and relations among the different dimensional entries. Experimental results show that SimulE can effectively model complex relation types and achieve state-of-the-art performance in almost all metrics on three benchmark datasets. In particular, compared with baseline ConvE, SimulE outperforms it in MRR by 2.9%, 9.8% and 2.8% on FB15k-237, YAGO3-10 and DB100K respectively.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"38 1","pages":"624-629"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84399534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152761
Kai Chen, Xiangmao Chang, Jun Zhan, Yanchao Zhao
Narrowband Internet of Things (NB-IoT) is an low-power wide area network based on cellar architecture. The repetition scheme is a key solution to achieve enhanced coverage with low complexity in the uplink. However, the impact of the current repetition scheme on energy consumption and coverage performance of NB-IoT are still unclear. In this paper, we conduct field measurements of the repetition scheme in terms of energy efficiency. We find that most of repetition values configured by the eNodeB lead to non-optimal energy efficiency. Then we propose an adaptive repetition scheme based on a regression block delivery rate (BDR) model which can be derived from a theoretical model and a small number of measurements. We conduct simulations based on real-world measurement data. The results show that the proposed adaptive repetition scheme outperforms the default repetition scheme in both energy efficiency and data transmission rate.
{"title":"Measurement and Optimization of Repetition Scheme in NB-IoT Uplink","authors":"Kai Chen, Xiangmao Chang, Jun Zhan, Yanchao Zhao","doi":"10.1109/CSCWD57460.2023.10152761","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152761","url":null,"abstract":"Narrowband Internet of Things (NB-IoT) is an low-power wide area network based on cellar architecture. The repetition scheme is a key solution to achieve enhanced coverage with low complexity in the uplink. However, the impact of the current repetition scheme on energy consumption and coverage performance of NB-IoT are still unclear. In this paper, we conduct field measurements of the repetition scheme in terms of energy efficiency. We find that most of repetition values configured by the eNodeB lead to non-optimal energy efficiency. Then we propose an adaptive repetition scheme based on a regression block delivery rate (BDR) model which can be derived from a theoretical model and a small number of measurements. We conduct simulations based on real-world measurement data. The results show that the proposed adaptive repetition scheme outperforms the default repetition scheme in both energy efficiency and data transmission rate.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"47 1","pages":"1932-1937"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89948509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the data-driven economy, data sharing is a key approach to unleashing the true value of data. Blockchain, as a decentralized ledger, can provide a trusted data sharing platform in an untrusted environment. However, existing blockchain-based data sharing schemes suffer from inefficiency and inadequate protection of security and privacy. To address the above issues, we propose a blockchain-based privacy-preserving data sharing scheme with security-enhanced access control. In the scheme, a secure data sharing architecture using dual-blockchain and the interplanetary file system (IPFS) is presented to provide decentralized and scalable storage. Based on the architecture, a blockchain-assisted multi-authority attribute-based encryption (BA-MA-ABE) algorithm with efficient attribute revocation and computation is designed in our work. Our BA-MA-ABE lever-ages blockchain to securely manage partial decryption keys and provides fine-grained access control over encrypted data. We also devise smart contracts that can support traceable access control over the flow of data while protecting user identity privacy with verifiable attribute credentials. In comparison with some existing work, our scheme shows more comprehensive security features with lower user computation overhead.
{"title":"A Blockchain-Based Privacy-Preserving Data Sharing Scheme with Security-Enhanced Access Control","authors":"Benyu Li, Jing Yang, Yuxiang Wang, Xiaojun Huang, Junshuai Ren, Liming Wang","doi":"10.1109/CSCWD57460.2023.10152751","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152751","url":null,"abstract":"In the data-driven economy, data sharing is a key approach to unleashing the true value of data. Blockchain, as a decentralized ledger, can provide a trusted data sharing platform in an untrusted environment. However, existing blockchain-based data sharing schemes suffer from inefficiency and inadequate protection of security and privacy. To address the above issues, we propose a blockchain-based privacy-preserving data sharing scheme with security-enhanced access control. In the scheme, a secure data sharing architecture using dual-blockchain and the interplanetary file system (IPFS) is presented to provide decentralized and scalable storage. Based on the architecture, a blockchain-assisted multi-authority attribute-based encryption (BA-MA-ABE) algorithm with efficient attribute revocation and computation is designed in our work. Our BA-MA-ABE lever-ages blockchain to securely manage partial decryption keys and provides fine-grained access control over encrypted data. We also devise smart contracts that can support traceable access control over the flow of data while protecting user identity privacy with verifiable attribute credentials. In comparison with some existing work, our scheme shows more comprehensive security features with lower user computation overhead.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"29 1","pages":"825-830"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86644059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Deep Neural Network (DNN) training on a large scale is extremely time-consuming and computationally intensive, which is accelerated by distributed training. In recent years, pipeline parallelism has been developed, which enables partitioning the model across several devices, e.g. GPU, and training efficiency is improved by dividing data batches into micro-batches, with each of them processed by a different stage of the model. Currently, parallel training assumes pipeline placement and partitioning are static, with parameters updating each iteration, without accounting for freezing. This results in computational resources not being fully utilized. In this paper, we propose FreezePipe, a novel method for optimizing deep learning training that combines the freezing mechanism with pipeline parallel training. In FreezePipe, a lightweight method for determining the freezing strategy based on gradient changes is employed. Considering that resources need to be released based on the frozen layer, a lightweight model partitioning algorithm was designed to determine the optimal strategy for pipeline partitioning. Experimental results show that FreezePipe can reduce the training time by 64.5% compared to Torchgpipe on CIFAR-10 dataset without compromising any model performance.
{"title":"FreezePipe: An Efficient Dynamic Pipeline Parallel Approach Based on Freezing Mechanism for Distributed DNN Training","authors":"Caishan Weng, Zhiyang Shu, Zhengjia Xu, Jinghui Zhang, Junzhou Luo, Fang Dong, Peng Wang, Zhengang Wang","doi":"10.1109/CSCWD57460.2023.10152643","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152643","url":null,"abstract":"Deep Neural Network (DNN) training on a large scale is extremely time-consuming and computationally intensive, which is accelerated by distributed training. In recent years, pipeline parallelism has been developed, which enables partitioning the model across several devices, e.g. GPU, and training efficiency is improved by dividing data batches into micro-batches, with each of them processed by a different stage of the model. Currently, parallel training assumes pipeline placement and partitioning are static, with parameters updating each iteration, without accounting for freezing. This results in computational resources not being fully utilized. In this paper, we propose FreezePipe, a novel method for optimizing deep learning training that combines the freezing mechanism with pipeline parallel training. In FreezePipe, a lightweight method for determining the freezing strategy based on gradient changes is employed. Considering that resources need to be released based on the frozen layer, a lightweight model partitioning algorithm was designed to determine the optimal strategy for pipeline partitioning. Experimental results show that FreezePipe can reduce the training time by 64.5% compared to Torchgpipe on CIFAR-10 dataset without compromising any model performance.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"10 1","pages":"303-308"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82888753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With rapid development of mobile Internet, a great number of mobile applications has emerged, presenting a great explosion in mobile Internet traffic. Therefore, accurate classification of application traffic is necessary to more effectively manage mobile Internet traffic. However, the encryption of mobile application traffic gradually eliminates traditional classification approaches based on specific signatures, greatly increasing the difficulty of the classification of mobile application traffic. Therefore, we propose a novel multi-feature fusion (MFF)- based approach to enhance the accuracy of mobile application traffic classification. We also extract packet length sequence, byte sequence, statistical feature, etc. Then, we perform weighted fusions of features based on Relief-F algorithm to achieve the best set of features. Finally, we use machine learning techniques for application classification. Compared to several other feature extraction methods, MFF achieves an excellent performance with an accuracy of 97.6% for 16 mobile applications and a F1-score of over 99% for VPN-nonVPN.
{"title":"Multi-Feature Fusion Based Approach for Classifying Encrypted Mobile Application Traffic","authors":"Qingya Yang, Peipei Fu, Junzheng Shi, Bingxu Wang, Zhuguo Li, G. Xiong","doi":"10.1109/CSCWD57460.2023.10152687","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152687","url":null,"abstract":"With rapid development of mobile Internet, a great number of mobile applications has emerged, presenting a great explosion in mobile Internet traffic. Therefore, accurate classification of application traffic is necessary to more effectively manage mobile Internet traffic. However, the encryption of mobile application traffic gradually eliminates traditional classification approaches based on specific signatures, greatly increasing the difficulty of the classification of mobile application traffic. Therefore, we propose a novel multi-feature fusion (MFF)- based approach to enhance the accuracy of mobile application traffic classification. We also extract packet length sequence, byte sequence, statistical feature, etc. Then, we perform weighted fusions of features based on Relief-F algorithm to achieve the best set of features. Finally, we use machine learning techniques for application classification. Compared to several other feature extraction methods, MFF achieves an excellent performance with an accuracy of 97.6% for 16 mobile applications and a F1-score of over 99% for VPN-nonVPN.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"2004 1","pages":"1112-1117"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82963594","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152712
Xizhe Zhang, Xiong Zhang, Jiahao Hu, Yuesheng Zhu
Password authentication is a widely used identity authentication method for computer supported cooperative systems. However, the frequent occurrence of password leakage incidents has become a universal problem, and the leaked passwords seriously threaten the security of users’ unleaked passwords. In order to gain a deeper understanding of the relationship between users’ old passwords and new passwords and help users choose a securer new password when their old passwords are leaked, we propose a new targeted online guessing algorithm, Targuess-II+, based on old password in this article. As a new probabilistic algorithm, Targuess-II+ not only supports the application of strong transformation rules at any positions in a password, but also shows the transformation process from one password to another. Our analysis and experimental results have demonstrated that Targuess-II+ obtains better performance in terms of crack rate and efficiency compared with other existing algorithms.
密码认证是计算机支持的协作系统中广泛使用的一种身份认证方法。然而，密码泄露事件的频繁发生已经成为一个普遍存在的问题，泄露的密码严重威胁着用户未公开密码的安全。为了更深入地了解用户的旧密码和新密码之间的关系，帮助用户在旧密码泄露时选择更安全的新密码，本文提出了一种新的基于旧密码的针对性在线猜测算法targuuss - ii +。作为一种新的概率算法，Targuess-II+不仅支持在密码的任意位置上应用强变换规则，而且还显示了从一个密码到另一个密码的变换过程。我们的分析和实验结果表明，与其他现有算法相比，Targuess-II+在裂纹率和效率方面具有更好的性能。
{"title":"A New Targeted Online Password Guessing Algorithm Based on Old Password","authors":"Xizhe Zhang, Xiong Zhang, Jiahao Hu, Yuesheng Zhu","doi":"10.1109/CSCWD57460.2023.10152712","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152712","url":null,"abstract":"Password authentication is a widely used identity authentication method for computer supported cooperative systems. However, the frequent occurrence of password leakage incidents has become a universal problem, and the leaked passwords seriously threaten the security of users’ unleaked passwords. In order to gain a deeper understanding of the relationship between users’ old passwords and new passwords and help users choose a securer new password when their old passwords are leaked, we propose a new targeted online guessing algorithm, Targuess-II+, based on old password in this article. As a new probabilistic algorithm, Targuess-II+ not only supports the application of strong transformation rules at any positions in a password, but also shows the transformation process from one password to another. Our analysis and experimental results have demonstrated that Targuess-II+ obtains better performance in terms of crack rate and efficiency compared with other existing algorithms.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"15 1","pages":"1470-1475"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82994937","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152714
Jiaqi Liu, Kaiyu Huang, Lunchen Xie
Federated learning is a machine learning paradigm where many clients collaboratively train a machine learning model while ensuring the nondisclosure of local data sets. Existing federated learning methods conduct optimization over the same model structure, which ensures the convenience of parameter updates. However, the same structure among clients and the server may pose risks of privacy leakage as parameters from one’s model can fit in others’ models. In this paper, we propose a heterogeneous federated learning method to preserve privacy. Each client utilizes neural architecture search to determine distinct models via local data and update the server model via a federated learning framework with knowledge distillation. Besides, we develop a privacy-preserving binary low-rank matrix decomposition method (Blow), i.e., decomposing the output matrix into two low-rank binary matrices, to further ensure the secrecy of distilled information. A simple but efficient alternating optimization method is proposed to address a key subproblem arising from the binary low-rank matrix decomposition, which falls into the category of the Np-hard bipartite boolean quadratic programming. Based on extensive experiments over the image classification task, we show our algorithm provides satisfactory accuracy and outperforms baseline algorithms in both privacy protection and communication efficiency.
{"title":"HeteFed: Heterogeneous Federated Learning with Privacy-Preserving Binary Low-Rank Matrix Decomposition Method","authors":"Jiaqi Liu, Kaiyu Huang, Lunchen Xie","doi":"10.1109/CSCWD57460.2023.10152714","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152714","url":null,"abstract":"Federated learning is a machine learning paradigm where many clients collaboratively train a machine learning model while ensuring the nondisclosure of local data sets. Existing federated learning methods conduct optimization over the same model structure, which ensures the convenience of parameter updates. However, the same structure among clients and the server may pose risks of privacy leakage as parameters from one’s model can fit in others’ models. In this paper, we propose a heterogeneous federated learning method to preserve privacy. Each client utilizes neural architecture search to determine distinct models via local data and update the server model via a federated learning framework with knowledge distillation. Besides, we develop a privacy-preserving binary low-rank matrix decomposition method (Blow), i.e., decomposing the output matrix into two low-rank binary matrices, to further ensure the secrecy of distilled information. A simple but efficient alternating optimization method is proposed to address a key subproblem arising from the binary low-rank matrix decomposition, which falls into the category of the Np-hard bipartite boolean quadratic programming. Based on extensive experiments over the image classification task, we show our algorithm provides satisfactory accuracy and outperforms baseline algorithms in both privacy protection and communication efficiency.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"11 1","pages":"1238-1244"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88722861","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Image-based calculation of human pose similarity is one of the computer vision research fields. Most existing research uses the human skeleton joint to calculate the human pose similarity, but usually does not consider the influence of inaccurate recognition of skeleton joint on similarity calculation caused by the complex environment (such as the occlusion of body parts, etc.). We propose a human pose similarity calculation method based on partition weighted OKS model. Due to the influence of external factors such as occlusion, the skeleton joint extracted by the human pose estimation algorithm is inaccurate, which leads to the decrease of the accuracy of the human pose similarity calculation. We propose the partition rule of human skeleton joints and the dynamic strategy adjustment of partition weight. The partition weighted OKS model and a human pose similarity calculation method based on the partition weighted OKS model are given. The experimental results on datasets show that the proposed method for human pose similarity calculation is superior to the traditional one.
{"title":"A Human Pose Similarity Calculation Method Based on Partition Weighted OKS Model","authors":"Weihong Yang, Hua Dai, Haozhe Wu, Geng Yang, Meng Lu, Guineng Zheng","doi":"10.1109/CSCWD57460.2023.10152566","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152566","url":null,"abstract":"Image-based calculation of human pose similarity is one of the computer vision research fields. Most existing research uses the human skeleton joint to calculate the human pose similarity, but usually does not consider the influence of inaccurate recognition of skeleton joint on similarity calculation caused by the complex environment (such as the occlusion of body parts, etc.). We propose a human pose similarity calculation method based on partition weighted OKS model. Due to the influence of external factors such as occlusion, the skeleton joint extracted by the human pose estimation algorithm is inaccurate, which leads to the decrease of the accuracy of the human pose similarity calculation. We propose the partition rule of human skeleton joints and the dynamic strategy adjustment of partition weight. The partition weighted OKS model and a human pose similarity calculation method based on the partition weighted OKS model are given. The experimental results on datasets show that the proposed method for human pose similarity calculation is superior to the traditional one.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"222 1","pages":"1760-1765"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91164461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","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：