Pi Zero Watch w/ Real Time Captioning

I'm pretty much deaf and my hearing aid is failing to meet my needs. While the diagnostics and adjustments to the hearing aids are being made, I want a way to understand the people around me who don't know American Sign Language. Initially, I wanted to do a Augmented Reality headset but it was a little out of my budget range ($200+) and the watch approach was a good stepping stone to that goal.

Requirements:

• "Wake" word ("Craig") automatically starts transcribing speech

• "Wake" word ("Craig") uses tactile buzzer or light flash to get my attention

• Solid battery life of 2+ hours

• Quick translations/display of text

• Readable in sunlight

• Lightweight

• Python

Secondary Requirements:

• Python web server: for easier AR implementation later

• Detects dog barking or buzzers/alarms around house

• Focus on ease of transitioning to AR system - e.g. Apple's ARKit or some other Unity-based integration.

First thing to do is to acquire and organize your parts.

microphone -Current: https://www.amazon.com/Sony-ECMCS3-Omnidirectional-Stereo-Microphone/dp/B0058MJX4O/ref=sr_1_5?s=musical-instruments&ie=UTF8&qid=1496956788&sr=1-5&

We'll start with the software build. The reason being, it's far easier to correct the build with a HDMI monitor via Pi versus trying to SSH your way to fixing the problem.

The Software build:

Install Raspbian OS. The Raspberry Pi website is chock-full of useful introductions to the board and getting up and running.

• For the OS on raspberry pi's, I usually download onto my windows machine, have at least a 8GB microSD card on hand, dock with SD adapter and PC, then install with the etcher tool.

After updating and upgrading Raspbian OS packages, install the SpeechRecognition module from python's repo system:

• Open terminal on pi (enter one line then enter and let it install successfully until you have done each line):

sudo apt-get install portaudio19-dev python-all-dev
sudo apt-get install python-pyaudio python3-pyaudio
pip3 install PyAudio 
pip3 install SpeechRecognition

• After successful installation, run in terminal:

python3 -m speech_recognition

The beautiful thing about the Raspberry Pi Zero is that you are able to load up the recommended Raspbian OS and swap with other Pi Zeros.

Testing Requirements:

• Raspberry Pi W Boots and Displays default "Hello Worlds"

• Mobile battery function

• Microphone check function

https://github.com/PiSupply/PaPiRus

The Hardware build:

The most difficult part of this build will be soldering the display screen to the Raspberry Pi Zero W. I've had some practice with simple soldering and with cheaper components, so this will be a little bit of a stepping stone.

https://www.pi-supply.com/make/papirus-assembly-tips-and-gotchas/

https://github.com/PiSupply/PaPiRus/tree/master/hardware/PaPiRus%20Zero

http://raspi.tv/2016/playing-with-papirus-zero

SSH Test Screen:

from papirus import PapirusText
text = PapirusText([rotation = rot])
# Write text to the screen
# text.write(text)
text.write("hello world")