stt.md 2.8 KB

Speech To Text (STT)

This chapter describes how STT are working. This configuration must apply in the settings.yml.

The syntax used is YAML.

General defaults

The setting.yml defines the STT you want to use by default

default_speech_to_text: "type default STT here"

Then, still in the settings.yml file, each STT must set up its configuration following 'speech_to_text' tag :

speech_to_text:
   - STT1:
      STT1parameter1: "value option1"
      STT1parameter2: "value option2"
   - STT2:
      STT2parameter1: "value option1"

Current Available STT

Google

The google STT is based on the Google Speech Recognition API

parameter required default choices comments
key No None
language No en-US lang LCID string

Bing

The Bing STT is based on the Microsoft Bing Voice Recognition API

parameter required default choices comments
key YES None User info
language No en-US lang 7 languages

wit.ai

The wit.ai STT is based on the Microsoft Wit.ai API

parameter required default choices comments
key YES None User info
language No en-US lang

api.ai

The api.ai STT is based on the api.ai API

parameter required default choices comments
key YES None User info
language No en-US lang

Houndify

Not working yet ... in coming

parameter required default choices comments
key YES None User info
client_id YES None User info
language No en-US en-US

Full Example

In the settings.yml file :

default_speech_to_text: "google"
speech_to_text:
  - google:
      language: "fr-FR"
  - wit:
      key: "B5JI3YUSLYOYWNIZRNBVM34XUODME2K"
  - bing:
      key: "9e48dert65904838bedc11aea6b36fb0"
  - apiai:
      key: "e0cbff145af44944a6b9f82c0668b527"
      language: "fr"
  - houndify:
      key: "6ej90T7qAV74OYXk4X4vI2Xhk7wPsJu4aEZ0G5Ll-BMmV1JGtFpCxtSH9SmTY4G3bpEJ7a5y_GTQid-CAKI6vw=="
      client_id: "lM2JXeaSticbSo9-llczbA=="