Stress Speech Identification Using Various Neural Networks
Journal Title: International Journal of Engineering and Science Invention - Year 2018, Vol 7, Issue 5
Abstract
There has been a Speech Identification technology already available in the market based on different assumptions. The work presented in this paper investigates the feasibility of alternative approaches for solving the problem more efficiently. A Speech Identification particularly Stress component in the research comprised of three distinct blocks, a Feature Extraction Neural Networks and Identifier. The Feature Extractor block uses MFCC which is most prominent and efficient for Stress Speech Identifier. Designs of the Neural Network blocks based on four different approaches are compared. The performance of Support Vector Machine, Multi-Layer Perceptrons, Radial Basis Functions and Recurrent Neural Networks based Identifier is tested on Speech which consists of Stress components Identification problem. Experimental results indicate that applying these neural networks reduces the training complexity and the operation of the Identifier. During the implementation of this all algorithms some results have been obtained in terms of the accuracy and the quality of the Identifier. The comparison between the different approaches to the design of the Identifier conducted here gives a better understanding of the problem and its possible solutions for well being of the society.
Authors and Affiliations
Mrs. N. P. Dhole, Dr. S. N. Kale
Keywords
Related Articles
- EP ID EP397069
- DOI -
- Views 54
- Downloads 0
Evaluation of Optical Performance of Low-Mounted Road Lighting Using Probeam Light Distribution
Ultimateobjective of road lighting is to provide the drivers with visual information for safe driving at night and from such a standpoint, installing the road lighting and securing the visibility are the optimal solution...
Electron impact total ionization cross sections of BH2 , SH, F2O , HCN , C2 and C3 molecules.
Total ionization cross sections for BH2, F2O,SH , HCN ,C2 and C3 molecules are studied theoretically in the energy range from ionization threshold to 2 Kev . The total ionization cross sections are calculated by Binary E...
Innovation in Indian Software Industry: An Empirical analysis
Indian software industry’s meteoric rise in the global software market since the 1990s has attracted the attention of academics and corporations alike. While numerous domestic and global corporations have set up their un...
A Study on Tsunami Protection Works
Coastal communities around the world are at risk to tsunami. The study was piloted to idealize the efficacy of sea wall in controlling the tsunami forces on onshore structures. In front of building model different types...
Security Management perspective for Internet of Things
Internet of Things (IoT) is one kind of a network where each and every little real world objects will be connected to each other and to the Internet. This will give a huge advantage to the society however it will likewis...