Since I saw the first Iron Man film, I wanted my own JARVIS. I mean, who doesn’t? So I looked into trying to create my own JARVIS with a Raspberry Pi and came across a great project called Jasper which worked great on my Raspberry Pi Model B and B+, but I had great difficulty getting it to work on the Pi 2.
Eventually though, my patience paid off as I managed to get it working using Google STT and a cheap (~£5) USB audio device. Here’s my how to get it working;
- Download the image, burn it to an SD card. More details on burning Raspbian images onto SD Cards can be found here.
- Edit the standard Jasper profile.yml file to accommodate the relevant information (or just run `./jasper.py populate`).
This install uses the Google Speech API so you will need an API key (follow the original instructions on Jasper for this), but should work with other speech engines.
It’s worth noting that I originally had it setup using PocketSphinx but I found the quality of the recognition to be poor – hence my decision to use Google.