Mozilla Deepspeech for Ros via Jackaudio as part of a larger Pipeline.
DeepSpeech4Ros is in development and currently not recommended for use. Requirements will most likely change in the near future; We will move away from jackaudio as sound framework to a more dedicated framework called esiaf, which is currently beeing worked on..
-
Jack audio
-
ROS
-
Deepspeech (install script included)
-
Boost library
-
soxr library
-
SpeechRecPipelineMsgs (https://github.com/Slothologist/SpeechRecPipelineMsgs)
Deepspeech is a bit tricky to install, since we need a linkable library and not just a python wheel. Therefore we must build and install Deepspeech ourselves, along with its dependencies, tensorflow and bazel.
But worry not friend, a install script for ubuntu is included in this repo (install_deepspeech.sh) to take all the effort out of that. It will create shared libraries for deepspeech and copy them to the DeepSpeech4Ros/libs folder, where the CMakeLists.txt will be looking for it.
Also included is a script to install the necessary system dependencies for tensorflow and bazel (install_prerequisites.sh).
You will need to install Jack, Ros, Boost and soxr yourself.