Sample code illustrating how to connect MT and ASR engines to Pervoice intergration platform.
This directory contains an SDK for connecting engines such as for speech recognition or machine translation to PerVoice' service architecture in the Cloud (MCloud) as well as for implementing clients that can use the service architecture in order to perform speech-to-speech translation on some audio data.
doc Additional documentation that describes the API and its usage include Header files necessary for developing own Backends and Clients Linux Example Backend and Client applications with source code and pre-compiled binaries for 64-bit Linux
v0.5 First initial release v0.6 several improvements and bugfixes partly based on comments from PerVoice
Chiara Canton [email protected]
(c) 2013 PerVoice S.p.A.
===================================================================================================
Create audiorecorder and ebclient:
win32 (in visualstudio 2013 shell 32-bit): in windows folder nmake -f makefile.win --> creates exes in exe32 nmake -f makefile.win clean32 --> clean exe32 and .obj files nmake -f makefile.win all-binary--> creates exes (included client_binary and worker_binary) in exe32
win64 (in visualstudio 2013 shell 64-bit): in windows folder nmake -f makefile.win all64 --> creates exes in exe64 nmake -f makefile.win clean64 --> clean exe64 and .obj files nmake -f makefile.win all-binary64--> creates exes (included client_binary and worker_binary) in exe64
nmake -f makefile.win clean --> clean exe32 and exe64 and all .obj files
-- created audiorecorder.exe and ebclient.exe in exe32 and exe64 (created also client_binary.exe and worker_binary.exe if all-binary*) --
unix32: in linux folder
TARGET_PLATFORM=i686 make -f makefile.linux
export LD_LIBRARY_PATH=./lib32
unix64: in linux folder
TARGET_PLATFORM=x86_64 make -f makefile.linux
export LD_LIBRARY_PATH=./lib64
-- created audiorecorder and ebclient in linux directory --