No abstract

No abstract

Abstract.A novel segmentation algorithm for brain images is proposed using finite skew Gaussian mixture model. Recently, much work has been reported in medical image segmentation. Among these techniques, finite Gaussian mixture models are considered to be more recent and accurate. However, in this approach, a number of segments that an image can be divided are taken through apriori and if these segments are not initiated properly it leads to misclassification. Hence, to overcome this disadvantage, we proposed an algorithm for Medical Image Segmentation using Hierarchical Clustering and Skew Gaussian Mixture. The experimentation is done with four different brain images and the results obtained are evaluated using Quality metrics.

Today major section of automatic speaker verification (ASV) research is focused on multiple objectives like optimization of feature subset and minimization of Equal Error Rate (EER). As such, numerous systems for feature dimension reduction are proposed. This includes framework coaching and testing analysis for every feature set that could be a time esurient trip. Because of its significance, the issue of feature selection has been researched by numerous scientists. In this paper, a new feature subset selection procedure is presented. Hybrid of Ant Colony and Artificial Bee Colony optimized the feature subset over 85% thereby decreased the computational complexity of ASV. Additionally an external record is maintained to store non-dominated solution vectors for which concept of Pareto dominance is used. An overall optimization of 87% is achieved thereby improved the recognition rate of ASV.

In this paper, we present an efficient speaker identification system based on generalized gamma distribution. This system comprises of three basic operations, namely speech features classification and metrics for evaluation. The features extracted using MFCC are passed to shifted delta cepstral coefficients (SDC) and then applied to linear predictive coefficients (LPC) to have effective recognition. To demonstrate our method, a database is generated with 200 speakers for training and around 50 speech samples for testing.Above 90% accuracy reported.