Pub Date : 2021-01-30DOI: 10.1109/DSC49826.2021.9346237
Yuan Li, Haiyan Wang, Shulan Wang, Yong Ding
With the widespread application of cloud computing technology, data privacy security problem becomes more serious. The recent studies related to searchable encryption (SE) area have shown that the data owners can share their private data with efficient search function and high-strength security. However, the search method has yet to be perfected, compared with the plaintext search mechanism. In this paper, based LSSS matrix, we give a new searchable algorithm, which is suitable for many search method, such as exact search, Boolean search and range search. In order to improve the search efficiency, the 0, 1-coding theory is introduced in the process of ciphertext search. Meanwhile it is shown that multi-search mechanism can improve the efficiency of data sharing. Finally, the performance analysis is presented, which prove our scheme is secure, efficient, and human-friendly.
{"title":"Attribute-Based Searchable Encryption Scheme Supporting Efficient Range Search in Cloud Computing","authors":"Yuan Li, Haiyan Wang, Shulan Wang, Yong Ding","doi":"10.1109/DSC49826.2021.9346237","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346237","url":null,"abstract":"With the widespread application of cloud computing technology, data privacy security problem becomes more serious. The recent studies related to searchable encryption (SE) area have shown that the data owners can share their private data with efficient search function and high-strength security. However, the search method has yet to be perfected, compared with the plaintext search mechanism. In this paper, based LSSS matrix, we give a new searchable algorithm, which is suitable for many search method, such as exact search, Boolean search and range search. In order to improve the search efficiency, the 0, 1-coding theory is introduced in the process of ciphertext search. Meanwhile it is shown that multi-search mechanism can improve the efficiency of data sharing. Finally, the performance analysis is presented, which prove our scheme is secure, efficient, and human-friendly.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121563626","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346245
Yi-Fan Tseng, Shih-Jie Gao
With the rapid development of Internet technologies, emerging network environments have been discussed, such as Internet of Things. In this manuscript, we proposed a novel subset predicate encryption for the access control in Internet of Things. Compared with the existing subset predicate encryption schemes, the proposed scheme enjoy the better efficiency due to the short private key and the efficient decryption procedure.
{"title":"Efficient Subset Predicate Encryption for Internet of Things","authors":"Yi-Fan Tseng, Shih-Jie Gao","doi":"10.1109/DSC49826.2021.9346245","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346245","url":null,"abstract":"With the rapid development of Internet technologies, emerging network environments have been discussed, such as Internet of Things. In this manuscript, we proposed a novel subset predicate encryption for the access control in Internet of Things. Compared with the existing subset predicate encryption schemes, the proposed scheme enjoy the better efficiency due to the short private key and the efficient decryption procedure.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"149 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122415340","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346241
Kui Zhu, Yongjun Ren, Qingfeng Zhu
To maintain the availability and durability of data, data owners (DOs) usually store their mass data in remote cloud, which is namely called cloud storage. Cloud storage is one of the most important applications of cloud computing. Through the technique of cloud storage, DOs can enjoy the services and benefits of cloud computing. However, while bringing the convenience and efficiency, there must be some security issues like data corrupted threatened in the process of using cloud storage services. Through many provable data possession (PDP) protocols has been presented to deal with these threats, most of them don not consider the problem that how should the protocol proceed when the corrputed data was found. In this paper, we propose a provable data possession protocol with fault tolerance including corrupted data locating and recovering utilizing Cuckoo Filter and Reed-Solomon codes respectively. Finally, we illustrate the security and performance of the proposed schemes, which shows the practicability of the proposed protocol.
{"title":"A Provable Data Possession Protocol in Cloud Storage Systems with Fault Tolerance","authors":"Kui Zhu, Yongjun Ren, Qingfeng Zhu","doi":"10.1109/DSC49826.2021.9346241","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346241","url":null,"abstract":"To maintain the availability and durability of data, data owners (DOs) usually store their mass data in remote cloud, which is namely called cloud storage. Cloud storage is one of the most important applications of cloud computing. Through the technique of cloud storage, DOs can enjoy the services and benefits of cloud computing. However, while bringing the convenience and efficiency, there must be some security issues like data corrupted threatened in the process of using cloud storage services. Through many provable data possession (PDP) protocols has been presented to deal with these threats, most of them don not consider the problem that how should the protocol proceed when the corrputed data was found. In this paper, we propose a provable data possession protocol with fault tolerance including corrupted data locating and recovering utilizing Cuckoo Filter and Reed-Solomon codes respectively. Finally, we illustrate the security and performance of the proposed schemes, which shows the practicability of the proposed protocol.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114417939","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346262
Haoyang An, Jiageng Chen
Recently, Onscreen Marking (OSM) system based on the traditional cloud platform has been widely used in various large-scale public examinations. However, the mainstream examination marking process is not transparent, and there is the possibility of a black-box operation, which damages the fairness of the examination. Also, the issues related to data security and privacy are still considered to be serious challenges. In this paper, we deal with the above problems by providing secure and private transactions in a distributed OSM assuming the semi-trusted examination center. We have implemented a proof-of-concept for a consortium blockchain-based OSM in a privacy-preserving and auditable manner, enabling markers to anonymously mark to the distributed ledger,
{"title":"ExamChain: A Privacy-Preserving Onscreen Marking System based on Consortium Blockchain","authors":"Haoyang An, Jiageng Chen","doi":"10.1109/DSC49826.2021.9346262","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346262","url":null,"abstract":"Recently, Onscreen Marking (OSM) system based on the traditional cloud platform has been widely used in various large-scale public examinations. However, the mainstream examination marking process is not transparent, and there is the possibility of a black-box operation, which damages the fairness of the examination. Also, the issues related to data security and privacy are still considered to be serious challenges. In this paper, we deal with the above problems by providing secure and private transactions in a distributed OSM assuming the semi-trusted examination center. We have implemented a proof-of-concept for a consortium blockchain-based OSM in a privacy-preserving and auditable manner, enabling markers to anonymously mark to the distributed ledger,","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115546709","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346277
Yi-Ming Chen, Chun-Hsien Yang, Guo-Chung Chen
In the field of mobile malware detection, the problem of sample imbalance often occurs in the dataset, making the classifier unable to learn features through sufficient data during the training process. This research used the generative adversarial networks (GAN). In this paper, features of malwares are transformed into image expressions, and data is generated from a small number of malicious families to balance and expand the original dataset. We also compare other data augmentation techniques to explore whether they are beneficial to identify a small number of malicious samples. Experiments show that both traditional techniques and GAN can improve the accuracy of classification, but GAN can more effectively improve the classification model to detect that the dataset originally has a small number of datasets and the recognition accuracy is lower. The experimental results show that in the different datasets of 4,000 data in Drebin and 20,000 data in AMD, the types with a relatively small number of samples are augmented by the GAN. Compared with before and after data augmentation, the difference in F1-score accuracy can reach 5%~20%.
{"title":"Using Generative Adversarial Networks for Data Augmentation in Android Malware Detection","authors":"Yi-Ming Chen, Chun-Hsien Yang, Guo-Chung Chen","doi":"10.1109/DSC49826.2021.9346277","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346277","url":null,"abstract":"In the field of mobile malware detection, the problem of sample imbalance often occurs in the dataset, making the classifier unable to learn features through sufficient data during the training process. This research used the generative adversarial networks (GAN). In this paper, features of malwares are transformed into image expressions, and data is generated from a small number of malicious families to balance and expand the original dataset. We also compare other data augmentation techniques to explore whether they are beneficial to identify a small number of malicious samples. Experiments show that both traditional techniques and GAN can improve the accuracy of classification, but GAN can more effectively improve the classification model to detect that the dataset originally has a small number of datasets and the recognition accuracy is lower. The experimental results show that in the different datasets of 4,000 data in Drebin and 20,000 data in AMD, the types with a relatively small number of samples are augmented by the GAN. Compared with before and after data augmentation, the difference in F1-score accuracy can reach 5%~20%.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130212596","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346240
Qian Mei, H. Xiong, Yanan Zhao, Kuo-Hui Yeh
By virtue of intelligent data processing and vehicular network, Internet of Vehicle (IoV) provides the transport system with real-time traffic information, which greatly improves the traffic conditions of the city. Moreover, to ease the computing and storage burden of the increasing number of the vehicle, edge computing is introduced to offload computing tasks on the local with low latency. But it still suffers from data integrity and privacy concerns. To meet these challenges, this paper proposes an efficient and conditional privacy-preserving authentication protocol using blockchain-enabled IoV with edge computing. Specifically, edge computing and consortium blockchains are combined to support efficient computation and storage capabilities with low communication delay while providing data auditability. Also, a pseudonym mechanism and identity-based signature are utilized to achieve identity privacy-preserving conditionally and messages authentication of vehicles, respectively. Moreover, for the dynamic change of vehicles, a clustering selection algorithm and a key updating algorithm based on Chinese remainder theorem are given, which guarantees the forward and backward security of transmission information. Security analysis and experiment evaluation with respect of communication and computation cost demonstrate the effectiveness of the proposed protocol.
车联网(Internet of Vehicle, IoV)通过智能数据处理和车联网，为交通系统提供实时交通信息，极大地改善了城市交通状况。此外，为了减轻日益增加的车辆数量带来的计算和存储负担，引入了边缘计算，将计算任务以低延迟的方式卸载到本地。但它仍然受到数据完整性和隐私问题的困扰。为了应对这些挑战，本文提出了一种高效的、有条件的隐私保护认证协议，该协议使用具有边缘计算的区块链支持的车联网。具体来说，边缘计算和联盟区块链相结合，以低通信延迟支持高效的计算和存储能力，同时提供数据可审计性。利用假名机制和基于身份的签名分别实现车辆的有条件身份隐私保护和消息认证。针对车辆的动态变化，给出了基于中国剩余定理的聚类选择算法和密钥更新算法，保证了传输信息的前向和后向安全性。从通信和计算成本两方面对协议进行了安全性分析和实验评估，验证了协议的有效性。
{"title":"Toward Blockchain-Enabled IoV with Edge Computing: Efficient and Privacy-Preserving Vehicular Communication and Dynamic Updating","authors":"Qian Mei, H. Xiong, Yanan Zhao, Kuo-Hui Yeh","doi":"10.1109/DSC49826.2021.9346240","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346240","url":null,"abstract":"By virtue of intelligent data processing and vehicular network, Internet of Vehicle (IoV) provides the transport system with real-time traffic information, which greatly improves the traffic conditions of the city. Moreover, to ease the computing and storage burden of the increasing number of the vehicle, edge computing is introduced to offload computing tasks on the local with low latency. But it still suffers from data integrity and privacy concerns. To meet these challenges, this paper proposes an efficient and conditional privacy-preserving authentication protocol using blockchain-enabled IoV with edge computing. Specifically, edge computing and consortium blockchains are combined to support efficient computation and storage capabilities with low communication delay while providing data auditability. Also, a pseudonym mechanism and identity-based signature are utilized to achieve identity privacy-preserving conditionally and messages authentication of vehicles, respectively. Moreover, for the dynamic change of vehicles, a clustering selection algorithm and a key updating algorithm based on Chinese remainder theorem are given, which guarantees the forward and backward security of transmission information. Security analysis and experiment evaluation with respect of communication and computation cost demonstrate the effectiveness of the proposed protocol.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122535689","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346272
S. S. Chaudhury, Sabyasachi Dutta, K. Sakurai
In this paper we propose a secret sharing scheme where the access structure changes over time following a dynamic data structure. The proposed scheme can accommodate an unbounded number of new parties and when the number of parties become very large, the scheme utilizes the dynamic data structure to search, add and delete parties which makes the scheme practical. Evolving Secret sharing was proposed in the literature to include unbounded number of parties into the scheme. We introduce a related idea called perpetual secret sharing which has the following advantages over the existing evolving schemes:- our construction reduces the memory usage of a dealer to a large extent and our scheme is implementable by constant-depth circuits $(AC^{0})$. Moreover, the construction is flexible and with some modifications, the construction can handle dynamic and evolving versions of several other multipartite access structures such as hierarchical access structures, compartmental access structures etc.
{"title":"Perpetual Secret Sharing from Dynamic Data Structures","authors":"S. S. Chaudhury, Sabyasachi Dutta, K. Sakurai","doi":"10.1109/DSC49826.2021.9346272","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346272","url":null,"abstract":"In this paper we propose a secret sharing scheme where the access structure changes over time following a dynamic data structure. The proposed scheme can accommodate an unbounded number of new parties and when the number of parties become very large, the scheme utilizes the dynamic data structure to search, add and delete parties which makes the scheme practical. Evolving Secret sharing was proposed in the literature to include unbounded number of parties into the scheme. We introduce a related idea called perpetual secret sharing which has the following advantages over the existing evolving schemes:- our construction reduces the memory usage of a dealer to a large extent and our scheme is implementable by constant-depth circuits $(AC^{0})$. Moreover, the construction is flexible and with some modifications, the construction can handle dynamic and evolving versions of several other multipartite access structures such as hierarchical access structures, compartmental access structures etc.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"5 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126117919","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346260
Xin Xie, Yu-Chi Chen
Blockchain has become very popular and suitable to the Internet of Things (IoT) field due to its non-tamperabaility and decentralization properties. The number of IoT devices and leaders (who own IoT devices) is increased exponentially, and thus data privacy and security are undoubtedly significant concerns. In this paper, we summarize some issues for the BeeKeeper system, a blockchain-based IoT system, proposed by Zhou et al., and then aim for presenting an improved solution for dencentralized data aggregation (DDA) on IoT. We propose our DDA system by using secret sharing to improve its efficiency, and smart contracts as the computing processors. In order to achieves data sharing (e.g., a leader to access data of others' devices), we use local differential privacy and cryptographic primitives such as token-based encryption. Finally, to show the feasibility, we provide some implementations and experiments for the DDA systems.
{"title":"Decentralized Data Aggregation: A New Secure Framework based on Lightweight Cryptographic Algorithms","authors":"Xin Xie, Yu-Chi Chen","doi":"10.1109/DSC49826.2021.9346260","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346260","url":null,"abstract":"Blockchain has become very popular and suitable to the Internet of Things (IoT) field due to its non-tamperabaility and decentralization properties. The number of IoT devices and leaders (who own IoT devices) is increased exponentially, and thus data privacy and security are undoubtedly significant concerns. In this paper, we summarize some issues for the BeeKeeper system, a blockchain-based IoT system, proposed by Zhou et al., and then aim for presenting an improved solution for dencentralized data aggregation (DDA) on IoT. We propose our DDA system by using secret sharing to improve its efficiency, and smart contracts as the computing processors. In order to achieves data sharing (e.g., a leader to access data of others' devices), we use local differential privacy and cryptographic primitives such as token-based encryption. Finally, to show the feasibility, we provide some implementations and experiments for the DDA systems.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127385960","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346271
K. Teo, T. BalamuraliB., Jer-Ming Chen, Jianying Zhou
Security for mobile devices have largely focused on the development of trusted hardware and securing software, however these secure platforms are still vulnerable to physical side channel attacks. Side channel attacks bypass secure hardware access controls, exploiting the physical characteristics of devices and onboard sensors to compromise and leak sensitive information. In this paper, we investigate the use of onboard sensors to recover user input on touchscreen interfaces. We evaluate the use of motion and acoustic sensors to categories user interactions with the device and apply machine learning techniques to find a strong correlation between acoustic emanations and user input. The acoustic output of a touch-screen mobile device is used to build a model that predicts user input with up to 86 % accuracy in a rpa listie scpnario_
{"title":"Retrieving Input from Touch Interfaces via Acoustic Emanations","authors":"K. Teo, T. BalamuraliB., Jer-Ming Chen, Jianying Zhou","doi":"10.1109/DSC49826.2021.9346271","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346271","url":null,"abstract":"Security for mobile devices have largely focused on the development of trusted hardware and securing software, however these secure platforms are still vulnerable to physical side channel attacks. Side channel attacks bypass secure hardware access controls, exploiting the physical characteristics of devices and onboard sensors to compromise and leak sensitive information. In this paper, we investigate the use of onboard sensors to recover user input on touchscreen interfaces. We evaluate the use of motion and acoustic sensors to categories user interactions with the device and apply machine learning techniques to find a strong correlation between acoustic emanations and user input. The acoustic output of a touch-screen mobile device is used to build a model that predicts user input with up to 86 % accuracy in a rpa listie scpnario_","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131993756","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 : 2021-01-30DOI: 10.1109/DSC49826.2021.9346266
Yan-Ting Liu, R. Chen, Christine Dewi
Convolutional Neural Networks (CNNs) achieve excellent in traffic sign detection and recognition with sufficient annotated training data. The quality of the whole vision system based on neural networks depend on the dataset. However, it is complicated to find traffic sign datasets from most of the countries of the world. In this context, Deep Convolutional Generative Adversarial Networks (DCGAN) can synthesize realistic and diverse additional training images to fill the data lack in the real image distribution. This paper mainly discusses how is the quality of the pictures generated by the DCGAN with various parameters. We use an image with a different number and size for training. Further, the Structural Similarity Index (SSIM) and MSE were used to evaluate the quality of the image. Our work measured SSIM values between generated images and corresponding real images. The generated images show high similarity with the actual image while using more images for training. The highest SSIM values reached when using 200 total images as input and images size 32×32.
{"title":"Generate Realistic Traffic Sign Image using Deep Convolutional Generative Adversarial Networks","authors":"Yan-Ting Liu, R. Chen, Christine Dewi","doi":"10.1109/DSC49826.2021.9346266","DOIUrl":"https://doi.org/10.1109/DSC49826.2021.9346266","url":null,"abstract":"Convolutional Neural Networks (CNNs) achieve excellent in traffic sign detection and recognition with sufficient annotated training data. The quality of the whole vision system based on neural networks depend on the dataset. However, it is complicated to find traffic sign datasets from most of the countries of the world. In this context, Deep Convolutional Generative Adversarial Networks (DCGAN) can synthesize realistic and diverse additional training images to fill the data lack in the real image distribution. This paper mainly discusses how is the quality of the pictures generated by the DCGAN with various parameters. We use an image with a different number and size for training. Further, the Structural Similarity Index (SSIM) and MSE were used to evaluate the quality of the image. Our work measured SSIM values between generated images and corresponding real images. The generated images show high similarity with the actual image while using more images for training. The highest SSIM values reached when using 200 total images as input and images size 32×32.","PeriodicalId":184504,"journal":{"name":"2021 IEEE Conference on Dependable and Secure Computing (DSC)","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132007824","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：