Software Speech Recognition
CSEF · 2003 Mathematics & Software
Overview
Objectives/Goals My project analyzed the performance and accuracy of different software libraries and how well they compared against each other. Methods/Materials Four different software libraries: Scansoft's Dragon Naturally Speaking Version 6 and Dragon Professional Solution Version 6, IBM Via Voice Version 9, and a written program designed to see how well it functions in contrast of the other software. Each software was tested in different texts: basic grammar, medical, legal documents, computer, and scientific terminology. Five tests for each software and for each text were completed and accuracy and performance was recorded. This information was placed in graphs and charts with the variables for percent correct and for the test number. A data sheet was made from features such as installation time, microphone, vocabulary storage, and etc. Results Via Voice contained fewer mistakes in all test categories with over 90% accuracy in basic grammar and scientific terminology and above 80% in the other three. Dragon Standard ,however, had above 85% accuracy in basic grammar and scientific terminology and above 79% accuracy in medical, legal documents, and computer. The testing was done to show that both Standard and Via Voice were comparable and each had its own advantages over the other. Dragon Professional was used to notice how close Standard and Via Voice were in accuracy rate. The home-designed software was created to compare how well software brought in store performs against one that is created at home. Conclusions/Discussion The information shows that even though Via Voice had the least mistakes, it was missing several key features such as a user-friendly wizard, or a proper method to help fix problems and answer common questions, which both of the Dragon software had. Looking at the price the written software was the cheapest of all the software. Via Voice was approximately half the price of Dragon Standard, which does put some serious thought into which software is the best to buy and use.
Summary statement
My project is about the analyzation, testing, and comparison of different software libraries.
Help received
Father taught how to write programs, helped prepare board, and made sure the testing was coducted properly.
Competition history
- CSEF 2003
Resources
Related projects
CSEF · 2004
Accuracy of Voice Recognition Software
CSEF · 2014
Did You Catch That?
CSEF · 2013
Misspelled! Creating an Accurate Computerized Spell Correcting Algorithm
CSEF · 2014
Creation of a Digital Voice Assistant Capable of Learning from Its Users
ISEF · 2018
Do You See What I'm Saying? A Voice to Text Human-Machine Interface for the Hearing Impaired
CSEF · 2023
Development of a Novel Automatic Speech Recognition to Reduce Racial and Gender Bias
CSEF · 2003
The Ability of Optical Character Recognition to Accurately Recognize and Identify Word-Processed Characters
CSEF · 2017
VibSpeech: A Communication Device for Both Visually and Aurally Impaired
Closest projects by meaning, across every fair and year in the corpus.
Browse more like this
Source: California Science & Engineering Fair public projects