use java speech api or matlab for writing a tts
I am working on a project about TTS. I have some modules written in matlab and some others in cpp. I should make an interface (like screen reader) and also use the matlab and cpp codes to prepare a complete TTS. I have just heard about java speech api and found it helpful, but to use this api I should convert th开发者_如何学Pythonese codes into java. Does anyone has any idea about choosing matlab or java for implementation, I mean java speech api can be better than matlab?
I appreciate any ideas.
Java Speech API doesn't technically do anything, as far as I can tell. It's an interface which other software may implement. I suggest using CMU Sphinx 4, which is an implementation of it. It comes with lots of demos and is easy to install, plus there are helpful forums and it's open source. I don't know what your project is, but if you would like to implement really low level stuff (like your own Fourier transform or something) you can implement the Sphinx DataProcessor API and plug it in without having to create an entire system yourself. Then you'd just put your class in the configuration file, similar to this:
<component name="frontEnd" type="edu.cmu.sphinx.frontend.FrontEnd">
<propertylist name="pipeline">
<item>microphone </item>
<item>premphasizer </item>
<item>windower </item>
<item>myfft </item>
<item>melFilterBank </item>
<item>dct </item>
<item>liveCMN </item>
<item>featureExtraction </item>
</propertylist>
</component>
<component name="myfft"
type="my.stuff.fastFourierTransform"/>
If you have existing matlab code you can convert it into java classes using this. If you have cpp code that you'd like to use, you can use the Java native interface.
精彩评论