[go: up one dir, main page]

Skip to content

respeaker/Alexa

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Alexa

New version is at https://github.com/respeaker/avs which uses the latest AVS HTTP/2 API, is more responsive and has more functions.

To use Alexe Voice Service with ReSpeaker.

Requirements

  • CherryPy
  • Requests
  • PyAudio
  • ReSpeaker python library
  • webrtcvad - for Voice Activity Detection, available on ReSpeaker by default
  • PocketSphinx - for Keyword Spotting, available on ReSpeaker
  • ffplay, part of ffmpeg

On ReSpeaker

Alexa will be installed at the lasest firmware of ReSpeaker. If the command alexa is available, skip step 1.

  1. Download alexa ipk and install it.
cd /tmp
wget https://github.com/respeaker/get_started_with_respeaker/raw/master/files/alexa_2017-01-18_ramips_24kec.ipk
opkg install alexa_2017-01-18_ramips_24kec.ipk
  1. Run alexa or /etc/init.d/alexa start to start Alexa Voice Service

  2. At the first time, you need to authorize the application.

Connect ReSpeaker's Access Point, go to http://192.168.100.1:3000 and tt will redirect you to Amazon to sign up or login in.

  1. Run alexa to interact with Alexa.

Note: if you get error IOError: [Errno -9998] Invalid number of channels, It's likely that mopidy-hallo or alexa is running and using the audio input channel. You can stop mopidy by running /etc/init.d/mopidy stop. /etc/init.d/mopidy disable will disable it to auto-run. /etc/init.d/alexa start will run alexa on background.

On Ubuntu

  1. Register for an Amazon Developer Account.

  2. Run git clone https://github.com/respeaker/Alexa.git && cd Alexa

  3. Rename example_creds.py to creds.py and fill ProductID, Security_Profile_Description, Security_Profile_ID, Client_ID and Client_Secret with your Alexa device information.

  4. Run sudo pip install cherrypy requests pyaudio webrtcvad pocketsphinx respeaker to get required python packages.

  5. You might also need these depdencies if you got errors at the above step: sudo apt-get install python-dev portaudio19-dev swig libpulse-dev. Then re-run step 4.

  6. Run python auth_web.py and open http://localhost:3000

    It will redirect you to Amazon to sign in. Make sure you have whitelisted the the above URL (with http:// not https:// in your app profile)

  7. Run python alexa.py to interact with Alexa.

Credits