Pub Date : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492797
H. M. Waidyasooriya, Daisuke Ono, M. Hariyama
Succinct data structures are introduced to efficiently solve a given problem while representing the data using as little space as possible. However, the full potential of the succinct data structures have not been utilized in software-based implementations due to the large storage size and the memory access bottleneck. This paper proposes a hardware-oriented data compression method to reduce the storage space without increasing the processing time. We use a parallel processing architecture to reduce the decompression overhead. According to the evaluation, we can compress the data by 37.5% and still have fast data access with small decompression overhead.
{"title":"Hardware-oriented succinct-data-structure based on block-size-constrained compression","authors":"H. M. Waidyasooriya, Daisuke Ono, M. Hariyama","doi":"10.1109/SOCPAR.2015.7492797","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492797","url":null,"abstract":"Succinct data structures are introduced to efficiently solve a given problem while representing the data using as little space as possible. However, the full potential of the succinct data structures have not been utilized in software-based implementations due to the large storage size and the memory access bottleneck. This paper proposes a hardware-oriented data compression method to reduce the storage space without increasing the processing time. We use a parallel processing architecture to reduce the decompression overhead. According to the evaluation, we can compress the data by 37.5% and still have fast data access with small decompression overhead.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134023547","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492790
Ghada Hamed, M. Marey, S. El-Sayed, M. Tolba
The information capacity is growing significantly as well as its level of importance and its transformation rate. In this paper, a blind data hiding hybrid technique is introduced using the concepts of cryptography and steganography in order to achieve double layer secured system. The proposed method consists of two phases: phase one is converting the message to DNA format using the proposed n-bits binary coding rule leading to high algorithm's cracking probability compared with those of other algorithms. Followed by applying the Playfair cipher based on DNA and amino acids to encrypt the secret message which generates ambiguity. Phase two is hiding the cipher secret message parts with the ambiguity results from from the first phase. The data is hidden using the least significant base (LSBase) only of each codon of a selected DNA reference sequence using 3:1 hiding strategy. The proposed technique achieves hiding the data in DNA with preserving its biological functions as possible without requiring any extra data to be sent to the receiver.
{"title":"Hybrid technique for steganography-based on DNA with n-bits binary coding rule","authors":"Ghada Hamed, M. Marey, S. El-Sayed, M. Tolba","doi":"10.1109/SOCPAR.2015.7492790","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492790","url":null,"abstract":"The information capacity is growing significantly as well as its level of importance and its transformation rate. In this paper, a blind data hiding hybrid technique is introduced using the concepts of cryptography and steganography in order to achieve double layer secured system. The proposed method consists of two phases: phase one is converting the message to DNA format using the proposed n-bits binary coding rule leading to high algorithm's cracking probability compared with those of other algorithms. Followed by applying the Playfair cipher based on DNA and amino acids to encrypt the secret message which generates ambiguity. Phase two is hiding the cipher secret message parts with the ambiguity results from from the first phase. The data is hidden using the least significant base (LSBase) only of each codon of a selected DNA reference sequence using 3:1 hiding strategy. The proposed technique achieves hiding the data in DNA with preserving its biological functions as possible without requiring any extra data to be sent to the receiver.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133736180","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492784
S. Miyamoto, Ryosuke Abe, Y. Endo, J. Takeshita
The Ward linkage method in agglomerative hierarchical clustering is sometimes used for non-Euclidean similarity, i.e., non-positive definite matrix of similarity, which is not an adequate use of this method, since the square Euclidean distance should be its basis. Nevertheless, this paper shows that the Ward method for non positive-definite similarity can partly be justified. It is shown that the result from the Ward method to a non positive-definite and normalized similarity is almost the same as another result from the Ward method to a positive-definite matrix obtained from the original similarity by adding a positive constant to the diagonal elements. More precisely, the same clusters are generated by the same order from the both data. Only the levels of their generations are different.
{"title":"Ward method of hierarchical clustering for non-Euclidean similarity measures","authors":"S. Miyamoto, Ryosuke Abe, Y. Endo, J. Takeshita","doi":"10.1109/SOCPAR.2015.7492784","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492784","url":null,"abstract":"The Ward linkage method in agglomerative hierarchical clustering is sometimes used for non-Euclidean similarity, i.e., non-positive definite matrix of similarity, which is not an adequate use of this method, since the square Euclidean distance should be its basis. Nevertheless, this paper shows that the Ward method for non positive-definite similarity can partly be justified. It is shown that the result from the Ward method to a non positive-definite and normalized similarity is almost the same as another result from the Ward method to a positive-definite matrix obtained from the original similarity by adding a positive constant to the diagonal elements. More precisely, the same clusters are generated by the same order from the both data. Only the levels of their generations are different.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115371418","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492817
T. Saitoh, Toshiki Shibata, Tsubasa Miyazono
We are studying image-based fish identification. Most of traditional approaches used a fish image which was easy to extract a fish region with a white background or uniform background for automatic processing. This research adapted an approach to give several points by manual operation by the user. The proposed approach is able to accept the fish image in the complicated background taken on the rocky place. Furthermore, to investigate the efficient features for fish recognition, we defined various features, such as, shape features, local features, and six kinds of texture features. We collected 129 species under various photography conditions, and the proposed method was carried out to it. As the results, it was confirmed that a combination features with geometric features and BoVW models obtained the highest recognition accuracy.
{"title":"Image-based fish recognition","authors":"T. Saitoh, Toshiki Shibata, Tsubasa Miyazono","doi":"10.1109/SOCPAR.2015.7492817","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492817","url":null,"abstract":"We are studying image-based fish identification. Most of traditional approaches used a fish image which was easy to extract a fish region with a white background or uniform background for automatic processing. This research adapted an approach to give several points by manual operation by the user. The proposed approach is able to accept the fish image in the complicated background taken on the rocky place. Furthermore, to investigate the efficient features for fish recognition, we defined various features, such as, shape features, local features, and six kinds of texture features. We collected 129 species under various photography conditions, and the proposed method was carried out to it. As the results, it was confirmed that a combination features with geometric features and BoVW models obtained the highest recognition accuracy.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"126 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115616869","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492799
Koki Kawasaki, T. Yoshikawa, T. Furuhashi
P300 speller is a system that allows users to input words using electroencephalogram (EEG). A component called P300 is used to interpret the EEG in P300 speller. In order to make a high performance P300 speller, it is essential to discriminate P300 from nonP300 precisely and automatically. In this study, deep learning (DL) is used to discriminate P300. The experimental result shows that DL was possible to discriminate P300 in EEG data, especially in the higher level layer. Furthermore, this study refers to the extracted feature by DL. We can see that DL learns feature from the waveforms correctly to discriminate P300 from others.
{"title":"Visualizing extracted feature by deep learning in P300 discrimination task","authors":"Koki Kawasaki, T. Yoshikawa, T. Furuhashi","doi":"10.1109/SOCPAR.2015.7492799","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492799","url":null,"abstract":"P300 speller is a system that allows users to input words using electroencephalogram (EEG). A component called P300 is used to interpret the EEG in P300 speller. In order to make a high performance P300 speller, it is essential to discriminate P300 from nonP300 precisely and automatically. In this study, deep learning (DL) is used to discriminate P300. The experimental result shows that DL was possible to discriminate P300 in EEG data, especially in the higher level layer. Furthermore, this study refers to the extracted feature by DL. We can see that DL learns feature from the waveforms correctly to discriminate P300 from others.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"380 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122794540","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492785
Dongming Tang
Serial analysis of gene expression (SAGE) is an efficient technique to produce a snapshot of the messenger RNA population in a sample. Clustering method has been widely used for SAGE data mining. In this study, we employ a new published measurement (maximal information coefficient, MIC) to measure the pair-wise correlation coefficients between SAGE libraries and then cluster together libraries with similar expression pattern. In addition, we present a clustering method named MicClustSAGE. We compared the results obtained by our method and hierarchical clustering with Pearson correlation. The experimental results exhibit the performance of the proposed method on several real-life SAGE datasets.
{"title":"Clustering analysis SAGE libraries using maximal information coefficient","authors":"Dongming Tang","doi":"10.1109/SOCPAR.2015.7492785","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492785","url":null,"abstract":"Serial analysis of gene expression (SAGE) is an efficient technique to produce a snapshot of the messenger RNA population in a sample. Clustering method has been widely used for SAGE data mining. In this study, we employ a new published measurement (maximal information coefficient, MIC) to measure the pair-wise correlation coefficients between SAGE libraries and then cluster together libraries with similar expression pattern. In addition, we present a clustering method named MicClustSAGE. We compared the results obtained by our method and hierarchical clustering with Pearson correlation. The experimental results exhibit the performance of the proposed method on several real-life SAGE datasets.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121001352","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492808
P. Hurtík, M. Vajgl, M. Burda
The paper introduces a real-life industrial problem: a jewelry stones classification. The stones are represented by their camera images. The goal of the contract was to evaluate stones into two (or more) specified classes according to their quality. Given requirements include very high processing speed and success rate of the classification. The goal of this paper is to publish a report of this contract and show a way how this task can be solved. In this paper we aim to usage of machine learning with respect to the image processing. We also design own learning and classification algorithm and answer the question if there is a place for a new machine learning algorithm. As an output of this paper a benchmark of the proposed algorithm with 81 state-of-the-art machine learning methods is presented.
{"title":"Jewelry stones classification: Case study","authors":"P. Hurtík, M. Vajgl, M. Burda","doi":"10.1109/SOCPAR.2015.7492808","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492808","url":null,"abstract":"The paper introduces a real-life industrial problem: a jewelry stones classification. The stones are represented by their camera images. The goal of the contract was to evaluate stones into two (or more) specified classes according to their quality. Given requirements include very high processing speed and success rate of the classification. The goal of this paper is to publish a report of this contract and show a way how this task can be solved. In this paper we aim to usage of machine learning with respect to the image processing. We also design own learning and classification algorithm and answer the question if there is a place for a new machine learning algorithm. As an output of this paper a benchmark of the proposed algorithm with 81 state-of-the-art machine learning methods is presented.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117351335","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492796
Haitham El-Hussieny, Samy F. M. Assal, A. Abouelsoud, S. M. Megahed, T. Ogasawara
In recent years, there has been an increasing interest in modeling natural human movements. The main question to be addressed is: what is the optimality criteria that human has optimized to achieve a certain movement. One of the most significant current discussions is the modeling of the reach-to-grasp movements that human naturally perform while approaching a certain object for grasping. Recent advances in Inverse Reinforcement Learning (IRL) approaches have facilitated investigation of reach-to-grasp movements in terms of the optimal control theory. IRL aims to learn the cost function that best describes the demonstrated human reach-to-grasp movements. Thus far, gradient-based techniques have been used to obtain the parameters of the underlying cost function. Such approaches, however, have failed to find the global optimal parameters since they are limited by locating only local optimum values. In this research, learning of the cost function for the reach-to-grasp movements is addressed as an Inverse Linear Quadratic Regulator (ILQR) problem, where linear dynamic equations and a quadratic cost are assumed. An efficient evolutionary optimization technique, Particle Swarm Optimization (PSO), is used to obtain the unknown cost for the reach-to-grasp movements under consideration. Moreover, an incremental-ILQR Algorithm is proposed to adjust the learned cost once new untrained demonstrations exist to overcome the over-fitting issue. The obtained results are encouraging and show harmony with those in neuroscience literature.
{"title":"Incremental learning of reach-to-grasp behavior: A PSO-based Inverse optimal control approach","authors":"Haitham El-Hussieny, Samy F. M. Assal, A. Abouelsoud, S. M. Megahed, T. Ogasawara","doi":"10.1109/SOCPAR.2015.7492796","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492796","url":null,"abstract":"In recent years, there has been an increasing interest in modeling natural human movements. The main question to be addressed is: what is the optimality criteria that human has optimized to achieve a certain movement. One of the most significant current discussions is the modeling of the reach-to-grasp movements that human naturally perform while approaching a certain object for grasping. Recent advances in Inverse Reinforcement Learning (IRL) approaches have facilitated investigation of reach-to-grasp movements in terms of the optimal control theory. IRL aims to learn the cost function that best describes the demonstrated human reach-to-grasp movements. Thus far, gradient-based techniques have been used to obtain the parameters of the underlying cost function. Such approaches, however, have failed to find the global optimal parameters since they are limited by locating only local optimum values. In this research, learning of the cost function for the reach-to-grasp movements is addressed as an Inverse Linear Quadratic Regulator (ILQR) problem, where linear dynamic equations and a quadratic cost are assumed. An efficient evolutionary optimization technique, Particle Swarm Optimization (PSO), is used to obtain the unknown cost for the reach-to-grasp movements under consideration. Moreover, an incremental-ILQR Algorithm is proposed to adjust the learned cost once new untrained demonstrations exist to overcome the over-fitting issue. The obtained results are encouraging and show harmony with those in neuroscience literature.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134051192","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492825
Sung-Bae Cho, Jun-Ho Kim
Conventional methods predict emotion directly by measuring equipment like electrode. However, this approach is not suitable for education, especially for children. In this paper, we propose modular Bayesian networks for predicting the emotion with the environment information from the sensors. The Bayesian network is constructed as modules divided by Markov boundary. To evaluate the proposed method, we use data collected from kindergarten classes. The results show more than 84% accuracy and 20 times faster than the single Bayesian network.
{"title":"Predicting group emotion in kindergarten classes by modular Bayesian networks","authors":"Sung-Bae Cho, Jun-Ho Kim","doi":"10.1109/SOCPAR.2015.7492825","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492825","url":null,"abstract":"Conventional methods predict emotion directly by measuring equipment like electrode. However, this approach is not suitable for education, especially for children. In this paper, we propose modular Bayesian networks for predicting the emotion with the environment information from the sensors. The Bayesian network is constructed as modules divided by Markov boundary. To evaluate the proposed method, we use data collected from kindergarten classes. The results show more than 84% accuracy and 20 times faster than the single Bayesian network.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123222116","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492766
Yih-Lon Lin, Chung-Ming Sung
In this paper, an approach of QR code detection using histograms of oriented gradients (HOG) and AdaBoost is proposed. There are two steps in our approach. In the first step, feature vectors are extracted using HOG with various cell sizes and overlapping or non-overlapping blocks. In the second step, the AdaBoost algorithms are trained by the input feature vectors from HOG and output targets. The QR code position is then detected via the predicted outputs from the AdaBoost algorithm. Experimental results show that the proposed method is an effective way to detect QR code position. Frankly speaking, the results reported here only provide preliminary study on QR code detection using HOG and AdaBoost.
{"title":"Preliminary study on QR code detection using HOG and AdaBoost","authors":"Yih-Lon Lin, Chung-Ming Sung","doi":"10.1109/SOCPAR.2015.7492766","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492766","url":null,"abstract":"In this paper, an approach of QR code detection using histograms of oriented gradients (HOG) and AdaBoost is proposed. There are two steps in our approach. In the first step, feature vectors are extracted using HOG with various cell sizes and overlapping or non-overlapping blocks. In the second step, the AdaBoost algorithms are trained by the input feature vectors from HOG and output targets. The QR code position is then detected via the predicted outputs from the AdaBoost algorithm. Experimental results show that the proposed method is an effective way to detect QR code position. Frankly speaking, the results reported here only provide preliminary study on QR code detection using HOG and AdaBoost.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115637323","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：