A Publisher & E-Learning Company
Client is app publishing company who wanted to build a language app to provide spoken language training to students. The challenge was to create a module to analyze the pronunciation of the student and match it with the standard pronunciation library. Phonetics and Voice Recognition is a complex subject and so no readily available tool or software was in place for the client to use in his application.
Our approach was to build this entire platform in small modules because the problem was very complex. Many times we had to go back to the beginning in order to re-build the modules from scratch. We researched and studied all available open source libraries for sound/voice editing and then selected the best for our purpose. The web interface to record spoken word was build by using Flex and the heavy lifting of Pronunciation analysis & Matching with library was done by using Java.
Java, JSP, Flex, AS3, XHTML, CSS
Results & Benefits
The technology platform developed by us is an innovative Pronunciation Analysis tool which can directly record the spoken word in a web browser and then can analyze the pronunciation of the user. It can match the pronunciation with standard library and can provide the correctness of the spoken word in percentages.