Classification based on Principal Component analysis has recently appeared in the literature in application to text-independent speaker identification. However, results have been reported for only clean speech data. I...
详细信息
Classification based on Principal Component analysis has recently appeared in the literature in application to text-independent speaker identification. However, results have been reported for only clean speech data. In this paper, we evaluate the performance of principal component classifier for text-independent speaker identification on telephone speech. We then improve its identification performance using a Vector Quantization classifier in combination, through fusion of classifier scores. An identification rate of 78.27% has been obtained on the NTIMIT database, which is well above the best identification rate ever reported in the literature obtained by using only one type of feature set.
Lately around the world the Information and Communications Technology (ICT) industry has taken interest in reducing the power consumption of ICT equipment. Communications service providers are especially interested in...
详细信息
In the recent years, rough set theory has been applied in diverse areas of research, however its application to classification problems is still a challenger. In this paper we present a new method to automatically gen...
详细信息
ISBN:
(纸本)9783642026102
In the recent years, rough set theory has been applied in diverse areas of research, however its application to classification problems is still a challenger. In this paper we present a new method to automatically generate fuzzy rules using an extension of rough sets. We use genetic algorithm to determine the granules of the knowledge to obtain the rough sets. The resulting classifier system based on the set of fuzzy rules was tested with the public databases: Iris, Wine, and Wdbc datasets, presenting accuracy rates of 100%, 100%, and 99%, respectively.
Based on moving least square, a multi-view car pose interpolation and corresponding recognition approach is proposed. This work firstly analyzes the shape characteristics of actual trace caused by ear pose varying in ...
详细信息
ISBN:
(纸本)9783642040191
Based on moving least square, a multi-view car pose interpolation and corresponding recognition approach is proposed. This work firstly analyzes the shape characteristics of actual trace caused by ear pose varying in feature space. Then according to training samples pose projection, we manage to recover the complete multi-view car pose manifold by using moving least square pose interpolation. The constructed multi-view ear pose manifolds can be easily utilized to recognize ear images captured under different views based on finding the minimal projection distance to the manifolds. The experimental results and some comparisons show the new method is superior to manifold learning method and B-Spline based recognition method.
Sample-based clustering is one of the most common methods for discovering disease subtypes as well as unknown taxonomies. By revealing hidden structures in microarray data, cluster analysis can potentially lead to mor...
详细信息
ISBN:
(纸本)9783642024801
Sample-based clustering is one of the most common methods for discovering disease subtypes as well as unknown taxonomies. By revealing hidden structures in microarray data, cluster analysis can potentially lead to more tailored therapies for patients as well as better diagnostic procedures. In this work, we present a novel method for automatically discovering clusters of samples which are coherent from a genetic point of view. Each possible cluster is characterized by a fuzzy pattern which maintains a fuzzy discretization of relevant gene expression values. Noise genes are identified and removed from the fuzzy pattern based on their probability of appearance. Possible clusters are randomly constructed and iteratively refined by following a probabilistic search and an optimization schema. Experimental results over publicly available microarray data show the effectiveness of the proposed method.
Sports coaches today have an access to a wide variety of information sources that describe the performance of their players. Cricket match data is highly available and rapidly growing in size which far exceeds the hum...
详细信息
ISBN:
(纸本)9781424429271
Sports coaches today have an access to a wide variety of information sources that describe the performance of their players. Cricket match data is highly available and rapidly growing in size which far exceeds the human abilities to analyze. Our major intention is to model an automated framework to identify specifics and correlations among play patterns, so as to haul out knowledge which can further be represented in the form of useful information in relevance to modify or improve coaching strategies and methodologies to confine performance enrichment at team level as well as individual. With this information, a coach can assess the effectiveness of certain coaching decisions and formulate game strategy for subsequent games. Since real time cricket data is too complex, Object-relational model is used to employ more sophisticated structure to store such data. Frequent pattern evaluation is imperative for sports be fond of cricket match data which facilitates recognition of main factors accounting for variances in data. While using simple apriori for interrelationship analysis, it is less time efficient because the raw data set which is too large and complex. On integrating association mining with Principal Component Analysis, the efficiency of mining algorithm is improved provided that Principal Component Analysis generates frequent patterns through statistical analysis and summarization not by repeated searching like other frequent patterns generation techniques. As the size and dimension of annotation database is large, Principal Component Analysis proceeds as a compression mechanism. Then the frequent patterns are analyzed for their interrelationship in order to generate interesting and confident rules of association.
Geometric moment invariant (GMI) is well known approach in patternrecognition. One of the weaknesses of GMI is in its invarianceness, where data or points concentrated near to the center-of-mass are neglected because...
详细信息
ISBN:
(纸本)9781424441549
Geometric moment invariant (GMI) is well known approach in patternrecognition. One of the weaknesses of GMI is in its invarianceness, where data or points concentrated near to the center-of-mass are neglected because of the existence of data or points that are far away from the center-of-mass. To solve this problem, Balslev *** has modified GMI method by adding a weighting function into GMIpsilas formula; thus we called it as Weighted Central Moment (WCM). WCM can increase noise tolerance for rotation/translation independent patternrecognition. In this paper, we present simulation results for characters with adjustable parameter alpha equal to 2/Rg. The experiments reveal that WCM yields intra-class results for identifying picture with different orientations. It also illustrates better inter-class distances in recognizing letter ldquogrdquo and ldquoqrdquo compared to GMI method.
The proceedings contain 62 papers. The topics discussed include: intelligent detection of LSB stego anomalies in images using softcomputing paradigms;human recognition using fusion of IRIS and ear data;an efficient a...
ISBN:
(纸本)9789380043579
The proceedings contain 62 papers. The topics discussed include: intelligent detection of LSB stego anomalies in images using softcomputing paradigms;human recognition using fusion of IRIS and ear data;an efficient approach for color image retrieval using Haar wavelet;swarm intelligence stability based on stochastic diffusion search;delta compression effect on smart card fingerprint verification system;an object-oriented conceptual level design for web data model;performance analysis of block matching criterion in video data on embedded processor using VHDL;medical image thresholding scheme using Atanassov intuitionistic fuzzy set;on locking conditions in m-sequence generators for the use in digital watermarking;performance simulation of Tahoe, Reno, New Reno and SACK over terrestrial and geostationary satellite links;and cryptanalysis of password authentication schemes: current status and key issues.
Scientific documents are unstructured data consisting of natural language and hard for scientists to read and manage. Keywords are very helpful for scientists to search the related documents and know about their conte...
详细信息
Faults of sensor data will always present in sensor networks because of unreliable communication links, measurement interference and harsh environment. Developing fusion algorithms that can tolerate faults is necessar...
详细信息
ISBN:
(纸本)9781849191388
Faults of sensor data will always present in sensor networks because of unreliable communication links, measurement interference and harsh environment. Developing fusion algorithms that can tolerate faults is necessary for reliable sensor network applications. In this paper, we study the fault tolerant fusion for moving vehicle classification based on Marzullo's interval fusion algorithm. The unreliable sensor data are represented using interval estimations. To reduce communication cost, quantized interval representation is adopted. Simulation results demonstrate the validity of the interval fusion algorithm. By using quantized representation, the communication cost is reduced.
暂无评论