The audio api lives on in the official audio api of eclipse smart home. The rest of the project currently is dormant and will no longer be maintained.
An experimental implementation of a voice command api for eclipse smarthome / openhab2 which defines:
- an audio api
- interfaces and sample implementations for voice to text and text to voice
- an interface and a sample implementaiton for a text command interpreter
- a webservice api for utilizing these to issue voice commands (speech to text to action to text to speech)
- clients for web and commandline to use this service
Current parts:
-
org.sylvani.io Core API for audio, voice2text, text2voice and text commands
Webservice API for audio & text based commands -
org.openhab.binding.sylvani.audio
Implementation of AudioSource and AudioOutput Interfaces for local audio system based on java sound api -
org.sylvani.client
Command line client for Websservice API -
org.sylvani.hal Sample implementation of CommandInterpreter (currently static mapped) Possible implementations could either build an fulltext index or a nlp model from the things and channels in the bus to optimize output
-
org.sylvani.mary
Sample implementation of Recognizer based on mary TTS -
org.sylvani.openhab.ui
Openhab UI Plugin based on recoderJS as client of webservie -
org.sylvani.oxford Sample implementaitons of recognizer and syntheizer based on microsoft oxford
-
org.sylvani.sphinx
Sample implementation of recognizer based on sphinx