[Proposal] Voice Enabled Automated Software (vEase)



                                       ABSTRACT
Our project ‘vEase’aims to make software applications easy to use and interactive by allowing users to communicate with their natural language. We are using a free, open source, java based framework called Sphinx which facilitates the use of an acoustic model to understand nepali language and native English tone. Our main work is focused on rigorously training the Sphinx system to understand nepali language as accurately as possible by using a tool called SphinxTrain, by developing our own speech trained acoustic model. This process is complex and involves recording our own audio, transcribing it, extensively training the system and later using it in our application to properly recognize user commands. Next important step would be to make a user interface using Java’s swing framework so that the user can effectively perform various activities like checking email, playing video, songs, pictures and reading online news using his/her voice commands instead of keyboard and mouse clicking in navigation wizards. All of our programming will be done in Java. We aim to make this software speaker independent so that it will be useful not only for general user but also for people with disabilities. This will prove to be a milestone project in incorporating the core concepts of Human Computer Interaction (HCI) and Natural Language Processing (NLP) that will bring evolution in the way people interact with computers. Keywords: natural language, framework Sphinx, SphinxTrain, Java swing framework, Human Computer Interface, Natural Language Processing.

No comments:

Post a Comment

its cool