futurenorthants/smartspeaker

A simple smart speaker

The aim of this project is to showcase innovation by building a smart speaker that can connect to Norbert's Microsft QnA service to answer questions.

Prerequisites

• Python 3.* installed
• Python virtual environments able to be created
• flac install (on mac with brew run brew install flac in ubuntu run sudo apt install flac)

Add a credentials.py file with the following format:

class credentials():
    def __init__(self):
        self.QnA_DOMAIN=""
        self.QnA_AUTH=""
        self.QnA_COOKIE=""
        self.QnA_ENDPOINT=""
        self.QnA_CONFIDENCE=60

Setup

Navigate to the folder connect to the virtual environment (on a mac run source virtenv/bin/activate, in vscode select python next to the git branch and select the recommended option) install the dependancies by running pip3 install -r requirements.txt

Run

ensure you are in the virtual environment run python3 app.py

smartspeaker's Issues

Optimise voice messages

Use a small file to generate the standard "how can I help" and "looking for an answer" into seperate files.

play those files in the main app.py rather than generating them every time

this will speed up the process as the text won't need sending to Google over the internet and no need to download the audio files.

Add internet check at start

pre-generate an audio file warning about no internet

do a check for internet access at the start of the app.py
if internet connection exists:
continue with application
if no internet connection:
play no internet audio file and end application