Standalone Gnuspeech Tube Resonance Model (TRM)

This repository contains a standalone version of the Tube Resonance Model (TRM) used in the Gnuspeech articulatory speech synthesizer. The code comes with a thin C wrapper that allows to control all aspects of the TRM in realtime. Furthermore, a Python wrapper is included that enables integration of TRM into any Python application.

How to build

Building the code was tested on Linux only. You'll need a C++11 compatible compiler. The code has no dependencies apart from the C++ standard library.

To compile the library, simply run

git clone https://github.com/astoeckel/gnuspeech_trm
cd gnuspeech_trm
make

In case the Makefile does not work for you, you can compile the executable by running

c++ -O3 -s gnuspeech_trm/*.cpp gnuspeech_trm/trm/*.cpp -o build/gnuspeech_trm_main

and the library by running

c++ -O3 -s -fPIC -shared gnuspeech_trm/trm/*.cpp -o build/libgnuspeech_trm.so

To test the code, run

build/gnuspeech_trm_main | aplay -f FLOAT_LE -c 1 -r 44100

To install the corresponding Python wrapper, run

pip3 install --user -e .

Test it by running

gnuspeech_trm/gnuspeech_trm.py | aplay -f FLOAT_LE -c 1 -r 44100

How to use

Please have a look at the code at the end of gnuspeech_trm/gnuspeech_trm.py for the Python wrapper and at gnuspeech_trm/gnuspeech_trm_main.cpp for the for the C wrapper. More details can as well be found in the C header gnuspeech_trm/gnuspeech_trm_main.h.

License

The code in the trm subfolder is licensed under the GPLv3 by various authors. The C and Python wrappers are (C) 2018 Andreas Stöckel, licensed under the GPLv3.