Early Sunday morning, we just decided to check if we could just add proper HID functionality with a few lines of new code. The main part was already working. Lucca suggested to send the characters selected directly (and not wait for entire words, seperated by a space) - so we did just that. It's even better, because you can go back further and delete mistakes, not limited to the current word.
Here is a little impression of it:

What's next?
- check dictionary letter by letter to map the frequency of letter combinations. We are working on the Python code for it (it millions of words to be checked character by character) ;-) Will be an interesting speed test for the computer :-))
- build an eyetracker that just distinguishes between up/down/left/right
- check back with professionals on improving and extending the device. Elias' grandnice has her own
Physiotherapy Institute and is interested in the project. She offered us help as soon as we got some proper working prototypes... [Thanks, Renata :-) !]
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.