Inspiration
We originally wanted to create a project using the OpenMV Cam M7. We've never used it before so we wanted to experiment with it by building something "useful". We decided on statically mounting the camera and having it be used to trigger a program that interfaces with Amazon Alexa's API
What it does
When a face is detected by the OpenMV Cam M7, it will forward a signal via GPIO to a raspberry pi. The raspberry pi has an on board client/server that negotiates requests with Amazon Alexa's API. Once the raspberry pi receives the signal from the OpenMV, it will trigger the microphone to start listening to speech. Once the end of a speech segment is detected, it will forward the response via the Alexa API to Amazon's servers to handle the parsing of the speech. The request is returned and the raspberry pi will output the answer to the query.
How I built it
We built a micro-python script which was compiled onto the openmv board that detects faces. We hooked the device to the raspberry pi via GPIO. We then modified the alexa-avs-sample application to suit our use case on the raspberry pi. We wrote automation scripts to facilitate the build and reliability of the program's life cycle.
Challenges I ran into
Maintaining a wireless connection on the raspberry pi is difficult.
Accomplishments that I'm proud of
What I learned
OpenMV is great!
What's next for Alexis
We plan to reintegrate the original voice cue to the application so that it ends up being a straight upgrade to the original Amazon Alexa.
Log in or sign up for Devpost to join the conversation.