- JavaME prototype client to record speaking voice.
- Sound file uploaded to server which will speech recognize sound into text.
- Text is translated using google translate API
- Text is then syntethised into sound file again.
- Sound file should be sent back to JavaME klient.
- JavaME client will playback the sound file.
## Deliverables
**Server should run in Java JRE 6.0 environment.
Client should run in JavaME environment.**
**Client - Runtime JavaME**
- Choose lang1 and lang2 (eg. English/Spanish or Spanish/English)
- Record spoken language from microphone
- Send sound file to server
- Recive sound file from server
- Playback sound file
**Server - Runtime Java JRE 6.0** (could be other)
- Accept incoming sound file from JavaME-client (wav?)
- Accept language variables from JavaME-client (lang1, lang2)
- Speech recognition wav file into text (lang1). Use Spinx-4 or similar for this recognition
([login to view URL])
- Translate lang1-text to lang2-text via Java API using Goole translate
([login to view URL])
- Convert lang2 text into speech-audio using FreeTTS 1.2 or similar.
([login to view URL])
- Send sound file to phone (wav?)
***
Server could have other specifications, development language/environment and tools.***
***
I am open to other suggestions and tools to use. But this prototype must work on the JavaME platform.***
Please check the PDF for graphical model.