Control the NeoPixels on the Seeed Studio ReSpeaker with your voice.
These instructions assume that you've set up the network on your ReSpeaker. See the Getting Started Guide for more info.
On your computer use Arduino IDE 1.6.11 to load lights.ino onto your ReSpeaker.
Clone the source code onto the SD Card of your ReSpeaker.
ssh to the ReSpeaker
ssh root@mylinkit.local
Clone this repo onto the SD card
cd /tmp/run/mountd/mmcblk0p1
git clone https://github.com/don/respeaker-lights.git
Libraries
cd respeaker-lights
pip install monotonic webrtcvad
Get a key for the Bing Speech API and create creds.py with the key
cp creds_template.py creds.py
vi creds.py # add the key and save
Run the app
python lights.py
Give the app time to start, then try "Turn the lights on", "Make the lights red", "Turn the brightness up", "Make the lights green", "Turn the lights off".
What's next? This app runs an open microphone and sends all text to Microsoft for voice to text, which isn't ideal. Ideally this could run in offline mode using pocketsphinx or use a combination of keyword spotting and Bing. See the pocketsphinx_keyword_spotting and respeaker_hi examples.
This code is based on the microsoft_cognitive_services and spi bridge ReSpeaker examples.