A GUI and debugging tool for the Speect TTS library.
DeSpeect is a graphical interface for Speect [Meraka Institute (2008-2013)], a library for the creation of vocal synthesis systems, which facilitates the inspection of its internal state during its execution and the writing of tests for its functionality.
- git (minimum version 2.14.1)
- curl ( minimum version 7.55.1)
- ANSI C/ISO C90 GCC compiler (minimum version 5.0)
- CMAKE (minimum version 2.8)
- SWIG (minimum Version 3.0.10)
- libxml2-dev (minimum version 2.9.4)
- python-dev (minimum version 2.7.14)
- Unix / Unix-like OS (DeSpeect has only been tested on Ubuntu 16.04 LTS)
- CMAKE (minimum version 2.8)
- ANSI C/ISO C90 GCC compiler (minimum version 7.0)
- Qt 5.9.0
You can run DeSpeect in two different ways, according to your needs:
- Developer build
- Clone or download the repository
- Execute the
build.sh
script in the root directory
This will install Speect, create a build for DeSpeect in DeSpeect/build/
and run the related tests.
- User build
- Clone or download the repository
- Execute the
Installer.sh
script in the root directory - Select a destination folder for the installation
This will install Speect and create a build for DeSpeect, without running any test.
The tests for DeSpeect are implemented through the Google Test framework and they are run at every build by Travis CI.
In order to run the tests locally, you need to perform a developer build which runs the tests automatically. Once you have installed the application, you can run the related tests by entering the build directory (DeSpeect/build/
) and using either the ctest
or unit_tests
commands.
You can run DeSpeect in two different ways, depending on how decided to install it:
- Developer build
- Enter the bin directory through the
DeSpeect/build/bin/
path - Start the application executing the
main
file
- User build
- Enter the root install directory
- Start the executable called
DeSpeect
In order to visualize a Speect internal graph, you just need to load a .json
voice file clicking on the Load Voice
button in the main window (sample voices are stored in the SpeectInstaller/voices/
directory), and subsequently execute Speect through the Run All
or Run Step
button. Notice that Run All
will execute all the utterance processors, while Run Step
will only run the first utterance processor (see [Speect docs](http://speect.sourceforge.net/ "Speect docs" for more information). The following is a detailed sample procedure of how to visualize a graph:
- Start the application
- Click on the
Load Voice
button - Open the file corresponding the
SpeectInstaller/voices/cmu_arctic_slt/voice.json
path (if you opted for a dev install) orDeSpeectInstall/Test/cmu_arctic_slt/voice.json
(if you opted for a user install) - Select the desired utterance processors in the homonymous sidebar
- Click the
Run step
button in order to execute the utterance processors list step by step, or clickRun all
in order to execute the whole list
- Qt 5.9 LTS
- CMAKE 2.8
- Speect 1.1.0-69-g65f4
- Ubuntu 16.04 LTS
- Travis CI