I need someone to create a voice recognition software that receives an mp3 or wav file and returns phoneme of what was said. It will have at maximum 3 words in each file. And needs to be able to receive a file with noise.
In this case the software will not try to recognize a word but write the phoneme that someone talk.
There is a lot of open source codes that tries to recognize a word, the job would be just extend and create a step-by-step how to install in a linux environment.
There is in Java a JavaSpeech library to work with.
Hope someone can easialy do this project.