ofxVosk is an openFrameworks addon of vosk-api to use automatic speech recognition (ASR) functionality based on Kaldi project.

Vosk is an open source speech recognition toolkit which is based on the Kaldi-ASR project. With this addon, you can add automatic speech recognition functionality to your openFrameworks app (which doesn't require an internet connection, supports various platform including RPi!)

License

Both Vosk and Kaldi are licensed under Apache-2.0 license. Since this project including those header files from vosk-api project, this project is also licensed under Apache-2.0 license.

Supported platform (so far)

• MacOSX 10.14 (Mojave), oF 0.11.0 (release)

How to use this?

OSX (example)

• You will get all the libraries and source code by cloning this repo from command line. git clone https://github.com/ryohajika/ofxVosk
• Go to VoskAPI's website and download one of these the models available on the website you want to use. (in the example project I used "vosk-model-small-en-us-0.3")
• Unpack the zipped model file and place it to examples/bin/data
• Run the project in Xcode