Pub Date : 2012-07-26DOI: 10.1109/ICCCNT.2012.6396061
V. Dandu, B. Ramkumar, H. Kittur
In this work we evaluated arrival profile of the HPM based multiplier partial products reduction tree in two ways: 1.manual delay, area calculation through logical effort, 2.ASIC implementation. Based on the arrival profile, we worked with some recently proposed optimal adders and finally we proposed an optimal hybrid adder for the final addition in HPM based parallel multiplier. This work derives some mathematical expressions to find the size of different regions in the partial product arrival profile which helps to design optimal adder for each region. This work evaluates the performance of proposed hybrid adder in terms of area, power and delay using 90nm technology. This work deals with manual calculation for 8-b and ASIC simulation of different adder designs for 8-b, 16-b, 32-b and 64-b multiplier bit sizes.
{"title":"Optimization of hybrid final adder for the high performance multiplier","authors":"V. Dandu, B. Ramkumar, H. Kittur","doi":"10.1109/ICCCNT.2012.6396061","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6396061","url":null,"abstract":"In this work we evaluated arrival profile of the HPM based multiplier partial products reduction tree in two ways: 1.manual delay, area calculation through logical effort, 2.ASIC implementation. Based on the arrival profile, we worked with some recently proposed optimal adders and finally we proposed an optimal hybrid adder for the final addition in HPM based parallel multiplier. This work derives some mathematical expressions to find the size of different regions in the partial product arrival profile which helps to design optimal adder for each region. This work evaluates the performance of proposed hybrid adder in terms of area, power and delay using 90nm technology. This work deals with manual calculation for 8-b and ASIC simulation of different adder designs for 8-b, 16-b, 32-b and 64-b multiplier bit sizes.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132613765","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6395945
T. Rao, S. Ramesh
This paper deals with the design of 2×2 Microstrip, Slotted rectangular waveguide antenna array and Patch-fed rod antennas needed for ultra high-speed, ultra high-capacity wireless communication systems which work at Millimeter waves at 60 GHz for different multimedia applications in indoor environments and illustrates its specifications and requirements. The antenna in such systems requires high gain, high-efficient, wider beamwidth and high performance design specifications. The designed antennas were simulated using electromagnetic software Agilent's EMPro and Ansys HFSS.
{"title":"Microstrip, slotted rectangular waveguide array and Patch-fed rod antenna design and simulation for Gigabit wireless communications at 60 GHz","authors":"T. Rao, S. Ramesh","doi":"10.1109/ICCCNT.2012.6395945","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6395945","url":null,"abstract":"This paper deals with the design of 2×2 Microstrip, Slotted rectangular waveguide antenna array and Patch-fed rod antennas needed for ultra high-speed, ultra high-capacity wireless communication systems which work at Millimeter waves at 60 GHz for different multimedia applications in indoor environments and illustrates its specifications and requirements. The antenna in such systems requires high gain, high-efficient, wider beamwidth and high performance design specifications. The designed antennas were simulated using electromagnetic software Agilent's EMPro and Ansys HFSS.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114901011","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6395967
S. Ravichandran, D. Sathya, R. Shanmugapriya, G. Isvariyaa
The main objective of this paper is to have a best association between customer and organisation. This method is proposed in order to discover knowledge from huge amount of data and to use the data efficiently because of great demand. Banking is the most commonly used application for financial section. In which, Enterprise Resource Planning (ERP) model is most widely used in order to cost control, accounting and e-business & analyses. The request of the customers are routed automatically to the next department when one department finishes their work of the customer's request and each department have access to the single database that holds the customers new request. Customer Relationship Management (CRM) model is responsible for receiving the request and sending responses to the customers quickly and directly. The request includes queries, complaints, suggestions, and orders. These requests are forwarded to inner view ERP through query generator. In this paper, we proposed a model that integrates the customer queries, transactions, databases and all other specifications used in ERP Systems, then use data mining techniques to integrate decision making and forecasting. Using ERP characteristics, data gathered from central database are in cluster format which is based on action taken against the queries generated by customers. Later the clustered data's are used by Apriori algorithm to extract new rules and patterns for the enhancement of an organisation.
{"title":"Rule-base data mining systems for customer queries","authors":"S. Ravichandran, D. Sathya, R. Shanmugapriya, G. Isvariyaa","doi":"10.1109/ICCCNT.2012.6395967","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6395967","url":null,"abstract":"The main objective of this paper is to have a best association between customer and organisation. This method is proposed in order to discover knowledge from huge amount of data and to use the data efficiently because of great demand. Banking is the most commonly used application for financial section. In which, Enterprise Resource Planning (ERP) model is most widely used in order to cost control, accounting and e-business & analyses. The request of the customers are routed automatically to the next department when one department finishes their work of the customer's request and each department have access to the single database that holds the customers new request. Customer Relationship Management (CRM) model is responsible for receiving the request and sending responses to the customers quickly and directly. The request includes queries, complaints, suggestions, and orders. These requests are forwarded to inner view ERP through query generator. In this paper, we proposed a model that integrates the customer queries, transactions, databases and all other specifications used in ERP Systems, then use data mining techniques to integrate decision making and forecasting. Using ERP characteristics, data gathered from central database are in cluster format which is based on action taken against the queries generated by customers. Later the clustered data's are used by Apriori algorithm to extract new rules and patterns for the enhancement of an organisation.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130718432","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6395992
N. Padmapriya, R. Rajmohan
Cloud computing is emanated as a useful paradigm for building applications with cloud services. Different types of resources can be published as services, and a many benefits are delivered with them. For the services to be more reliable to service consumers a reliability model should be proposed for cloud based services. Since in cloud many users' access the same information the services should have high reliability feature. Hence reliability is most important feature for cloud. In this paper, we propose a reliability evaluation suite for cloud services, which includes quality attributes and metrics. A case study is presented to show its applicability.
{"title":"Reliability evaluation suite for cloud services","authors":"N. Padmapriya, R. Rajmohan","doi":"10.1109/ICCCNT.2012.6395992","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6395992","url":null,"abstract":"Cloud computing is emanated as a useful paradigm for building applications with cloud services. Different types of resources can be published as services, and a many benefits are delivered with them. For the services to be more reliable to service consumers a reliability model should be proposed for cloud based services. Since in cloud many users' access the same information the services should have high reliability feature. Hence reliability is most important feature for cloud. In this paper, we propose a reliability evaluation suite for cloud services, which includes quality attributes and metrics. A case study is presented to show its applicability.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116881828","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6396005
R. Thiruvengadathan, S. Srikanth
In this paper we analyze the performance of different channel estimators under different antenna correlation conditions. This analysis is performed for various spatial multiplexing techniques considered in 3GPP Long term evolution (LTE) Release 8 (R8) using extensive simulations. By considering the LTE downlink system parameters and the LTE channel model, we measure the performance of channel estimators. Finally we present the performance of channel estimation on the performance of precoding in LTE systems.
{"title":"Performance of MIMO channel estimation in LTE downlink","authors":"R. Thiruvengadathan, S. Srikanth","doi":"10.1109/ICCCNT.2012.6396005","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6396005","url":null,"abstract":"In this paper we analyze the performance of different channel estimators under different antenna correlation conditions. This analysis is performed for various spatial multiplexing techniques considered in 3GPP Long term evolution (LTE) Release 8 (R8) using extensive simulations. By considering the LTE downlink system parameters and the LTE channel model, we measure the performance of channel estimators. Finally we present the performance of channel estimation on the performance of precoding in LTE systems.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131561390","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6396019
T. R. Sivapriya, A. Kamal, V. Thavavel
The objective of this study is to investigate the use of LSSVM (Least Square Support Vector Machine) trained with Chaotic PSO (Particle Swarm Optimization) for distinguishing different levels of Dementia from brain MRI. The availability of an effective method that is more objective than human readers can potentially lead to more reliable and reproducible dementia diagnostic procedures. The proposed scheme consists of several steps including feature extraction, feature selection and classification. This research paper proposes an intelligent classification technique to identify normal and demented patients using LSSVM. The manual interpretation of large volumes of brain MRI may lead to incomplete diagnosis. Hence the LSSVM approach is trained with multiple biomarkers to facilitate effective, accurate classification which is a requirement of the hour. SVM-PSO, LS-SVM-PSO classifiers are compared with LS-SVM trained by Chaotic PSO. LS-SVM-Chaotic PSO yields 100% accurate results and outperforms other classifiers in terms of sensitivity, specificity and accuracy in this analysis.
{"title":"Automated classification of MRI based on hybrid Least Square Support Vector Machine and Chaotic PSO","authors":"T. R. Sivapriya, A. Kamal, V. Thavavel","doi":"10.1109/ICCCNT.2012.6396019","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6396019","url":null,"abstract":"The objective of this study is to investigate the use of LSSVM (Least Square Support Vector Machine) trained with Chaotic PSO (Particle Swarm Optimization) for distinguishing different levels of Dementia from brain MRI. The availability of an effective method that is more objective than human readers can potentially lead to more reliable and reproducible dementia diagnostic procedures. The proposed scheme consists of several steps including feature extraction, feature selection and classification. This research paper proposes an intelligent classification technique to identify normal and demented patients using LSSVM. The manual interpretation of large volumes of brain MRI may lead to incomplete diagnosis. Hence the LSSVM approach is trained with multiple biomarkers to facilitate effective, accurate classification which is a requirement of the hour. SVM-PSO, LS-SVM-PSO classifiers are compared with LS-SVM trained by Chaotic PSO. LS-SVM-Chaotic PSO yields 100% accurate results and outperforms other classifiers in terms of sensitivity, specificity and accuracy in this analysis.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132434131","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6395998
S. P. Ngayarkanni, N. Kamal, V. Thavavel
Breast cancer is one of the most common forms of cancer in women. In order to reduce the death rate , early detection of cancerous regions in mammogram images is needed. The existing system is not so accurate and it is time consuming. The Proposed system is mainly used for automatic segmentation of the mammogram images and classifies them as benign, malignant or normal based on the decision tree ID3 algorithm. A hybrid method of data mining technique is used to predict the texture features which play a vital role in classification. The sensitivity, the specificity, positive prediction value and negative prediction value of the proposed algorithm were determined and compared with the existing algorithms. The size and the stages of the tumor is detected using the ellipsoid volume formula which is calculated over the segmented region. Automatic classification of the mammogram MRI images is done through three layered ANN .The weights are adjusted based on the rule extracted from ID3 algorithm .Both qualitative and quantitative methods are used to detect the accuracy of the proposed system.The sensitivity, the specificity, positive prediction value and negative prediction value of the proposed algorithm accounts to 99.78%, 99.9%, 94% and 98.5% which rates very high when compared to the existing algorithms. This paper focuses on the comparative analysis of the existing methods and the proposed technique in terms of sensitivity, specificity, accuracy, time consumption and ROC.
{"title":"Automatic detection and classification of cancerous masses in mammogram","authors":"S. P. Ngayarkanni, N. Kamal, V. Thavavel","doi":"10.1109/ICCCNT.2012.6395998","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6395998","url":null,"abstract":"Breast cancer is one of the most common forms of cancer in women. In order to reduce the death rate , early detection of cancerous regions in mammogram images is needed. The existing system is not so accurate and it is time consuming. The Proposed system is mainly used for automatic segmentation of the mammogram images and classifies them as benign, malignant or normal based on the decision tree ID3 algorithm. A hybrid method of data mining technique is used to predict the texture features which play a vital role in classification. The sensitivity, the specificity, positive prediction value and negative prediction value of the proposed algorithm were determined and compared with the existing algorithms. The size and the stages of the tumor is detected using the ellipsoid volume formula which is calculated over the segmented region. Automatic classification of the mammogram MRI images is done through three layered ANN .The weights are adjusted based on the rule extracted from ID3 algorithm .Both qualitative and quantitative methods are used to detect the accuracy of the proposed system.The sensitivity, the specificity, positive prediction value and negative prediction value of the proposed algorithm accounts to 99.78%, 99.9%, 94% and 98.5% which rates very high when compared to the existing algorithms. This paper focuses on the comparative analysis of the existing methods and the proposed technique in terms of sensitivity, specificity, accuracy, time consumption and ROC.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134366493","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6396075
R. Ramkumar, S. Arumugam
Goal of the proposed iris recognition is to recognize human identity through the textural characteristics of one's iris muscular patterns. Even though eye color is dependent on heredity, in contrast to this, iris is independent and uncorrelated even for twins. Out of various biometrics such as finger and hand geometry, face, ear and voice recognition, iris recognition has been acknowledged as one of the most accurate biometric modalities because of its high recognition rate. In this proposed iris recognition method, pupil localization is done by using negative function and four neighbours method so that irrespective of pupil's contour, either circle or ellipse, the pupil's boundary is detected accurately. For iris outer boundary detection, contrast enhancement, special wedges and thresholding techniques are used to isolate the specific iris regions without eyelid and eyelash occlusions. Now the resultant iris portion alone is transformed into polar coordinate system for normalization process. Histogram equalization technique is used for enhancing the normalized iris image. For feature extraction and matching process, cumulative sum-based change analysis and hamming distance are employed. When compared with the existing algorithms, this proposed algorithm is robust, accurate and also has low computational time and complexity.
{"title":"A novel iris recognition algorithm","authors":"R. Ramkumar, S. Arumugam","doi":"10.1109/ICCCNT.2012.6396075","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6396075","url":null,"abstract":"Goal of the proposed iris recognition is to recognize human identity through the textural characteristics of one's iris muscular patterns. Even though eye color is dependent on heredity, in contrast to this, iris is independent and uncorrelated even for twins. Out of various biometrics such as finger and hand geometry, face, ear and voice recognition, iris recognition has been acknowledged as one of the most accurate biometric modalities because of its high recognition rate. In this proposed iris recognition method, pupil localization is done by using negative function and four neighbours method so that irrespective of pupil's contour, either circle or ellipse, the pupil's boundary is detected accurately. For iris outer boundary detection, contrast enhancement, special wedges and thresholding techniques are used to isolate the specific iris regions without eyelid and eyelash occlusions. Now the resultant iris portion alone is transformed into polar coordinate system for normalization process. Histogram equalization technique is used for enhancing the normalized iris image. For feature extraction and matching process, cumulative sum-based change analysis and hamming distance are employed. When compared with the existing algorithms, this proposed algorithm is robust, accurate and also has low computational time and complexity.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134580317","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6395965
Du Ruiqing, Tan Jing, Zhou Ming
Magnetic data are often recorded in text files by magnetometers. They are not easy to deal with, and they are often transferred and written into Excel files to further process. The aim of this paper is to discuss how to use Visual C# to do the transferring work. Development tools, key technologies and flow chart of the program edited by Visual C# are introduced. The main steps of transferring work are elaborated.
{"title":"Transfer magnetic data from text files to excel files based on Visual C#","authors":"Du Ruiqing, Tan Jing, Zhou Ming","doi":"10.1109/ICCCNT.2012.6395965","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6395965","url":null,"abstract":"Magnetic data are often recorded in text files by magnetometers. They are not easy to deal with, and they are often transferred and written into Excel files to further process. The aim of this paper is to discuss how to use Visual C# to do the transferring work. Development tools, key technologies and flow chart of the program edited by Visual C# are introduced. The main steps of transferring work are elaborated.","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133241193","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 : 2012-07-26DOI: 10.1109/ICCCNT.2012.6477851
M. Abouelenien, Yiwen Wan, Abdullah N. Saudagar
Classification of actions by human actors from video enables new technologies in diverse areas such as surveillance and content-based retrieval. We propose and evaluate alternative models, one based on feature-level fusion and the second on decision-level fusion. Both models employ direct classification - inferring from low-level features the nature of the action. Interesting points are assumed to have salient 3D (spatial plus temporal) gradients that distinguish them from their neighborhoods. They are identified using three distinct 3D interesting-point detectors. Each detected interest point set is represented as a bag-of-words descriptor. The feature level fusion model concatenates descriptors subsequently used as input to a classifier. The decision level fusion uses an ensemble and majority voting scheme. Public data sets consisting of hundreds of action videos were used in testing. Within the test videos, multiple actors performed various actions including walking, running, jogging, handclapping, boxing, and waving. Performance comparison showed very high classification accuracy for both models with feature-level fusion having an edge. For feature-level fusion the novelty is the fused histogram of visual words derived from different sets of interesting points detected by different saliency detectors. For decision fusion besides Adaboost the majority voting scheme is also utilized in ensemble classifiers based on support vector machines, knearest neighbor, and decision trees. The main contribution, however, is the comparison between the models and, drilling down, the performance of different base classifiers, and different interest point detectors for human motion recognition
{"title":"Feature and decision level fusion for action recognition","authors":"M. Abouelenien, Yiwen Wan, Abdullah N. Saudagar","doi":"10.1109/ICCCNT.2012.6477851","DOIUrl":"https://doi.org/10.1109/ICCCNT.2012.6477851","url":null,"abstract":"Classification of actions by human actors from video enables new technologies in diverse areas such as surveillance and content-based retrieval. We propose and evaluate alternative models, one based on feature-level fusion and the second on decision-level fusion. Both models employ direct classification - inferring from low-level features the nature of the action. Interesting points are assumed to have salient 3D (spatial plus temporal) gradients that distinguish them from their neighborhoods. They are identified using three distinct 3D interesting-point detectors. Each detected interest point set is represented as a bag-of-words descriptor. The feature level fusion model concatenates descriptors subsequently used as input to a classifier. The decision level fusion uses an ensemble and majority voting scheme. Public data sets consisting of hundreds of action videos were used in testing. Within the test videos, multiple actors performed various actions including walking, running, jogging, handclapping, boxing, and waving. Performance comparison showed very high classification accuracy for both models with feature-level fusion having an edge. For feature-level fusion the novelty is the fused histogram of visual words derived from different sets of interesting points detected by different saliency detectors. For decision fusion besides Adaboost the majority voting scheme is also utilized in ensemble classifiers based on support vector machines, knearest neighbor, and decision trees. The main contribution, however, is the comparison between the models and, drilling down, the performance of different base classifiers, and different interest point detectors for human motion recognition","PeriodicalId":364589,"journal":{"name":"2012 Third International Conference on Computing, Communication and Networking Technologies (ICCCNT'12)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2012-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131699030","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：