Metadata-Version: 2.1
Name: neon-speech
Version: 3.2.2a2
Summary: Neon Speech Module
Home-page: https://github.com/NeonGeckoCom/neon_speech
Author: Neongecko
Author-email: developers@neon.ai
License: BSD-3-Clause
Description-Content-Type: text/markdown
License-File: LICENSE.md
Requires-Dist: ovos-core (~=0.0.7)
Requires-Dist: SpeechRecognition (~=3.8)
Requires-Dist: PyAudio (~=0.2)
Requires-Dist: mycroft-messagebus-client (~=0.10)
Requires-Dist: ovos-utils (~=0.0.30)
Requires-Dist: ovos-plugin-manager (~=0.0.19)
Requires-Dist: pydub (~=0.23)
Requires-Dist: click (~=8.0)
Requires-Dist: click-default-group (~=1.2)
Requires-Dist: neon-transformers (~=0.2)
Requires-Dist: neon-utils[audio] (~=1.3)
Requires-Dist: ovos-config (~=0.0.7)
Requires-Dist: ovos-vad-plugin-webrtcvad (~=0.0.1)
Requires-Dist: ovos-ww-plugin-vosk (~=0.1)
Provides-Extra: docker
Requires-Dist: ovos-stt-plugin-vosk (~=0.1) ; extra == 'docker'
Requires-Dist: neon-stt-plugin-nemo (~=0.0.2) ; extra == 'docker'
Requires-Dist: ovos-ww-plugin-pocketsphinx (~=0.1) ; extra == 'docker'
Requires-Dist: ovos-ww-plugin-precise-lite (~=0.1) ; extra == 'docker'
Requires-Dist: ovos-ww-plugin-precise (~=0.1) ; extra == 'docker'

# Neon Speech
Speech (Voice) Module for Neon Core. This module can be treated as a replacement for the
[mycroft-core](https://github.com/MycroftAI/mycroft-core) speech module. This module handles input audio, performs STT, 
and optionally returns the text or forwards it to the skills module for intent parsing.

## Neon Enhancements
`neon-speech` extends `mycroft-speech` with the following added functionality:
* Support for continuous STT without a wake word
* Audio processing module plugin system to modify audio and extract context
* Messagebus API listeners to handle outside requests for STT or inputs to the skills module
* Arbitrary configuration supported by passing at module init


## Compatibility
Mycroft STT and Wake Word plugins are compatible with `neon-speech`, with the exception of skipping wake words,
which is currently only supported by Neon STT plugins.

## Running in Docker
The included `Dockerfile` may be used to build a docker container for the neon_speech module. The below command may be used
to start the container.

```shell
docker run -d \
--network=host \
--name=neon_speech \
-v ~/.config/pulse/cookie:/tmp/pulse_cookie:ro \
-v ${XDG_RUNTIME_DIR}/pulse:${XDG_RUNTIME_DIR}/pulse:ro \
--device=/dev/snd:/dev/snd \
-e PULSE_SERVER=unix:${XDG_RUNTIME_DIR}/pulse/native \
-e PULSE_COOKIE=/tmp/pulse_cookie \
neon_speech
```
