Auto-correct images using google vision API?


The ability to auto-correct student images (show your work) using something like google vision API, would be useful. Perhaps have a definable threshold (%_match) of how well a student image matches the “key” image.
I had students draw lewis structure diagrams as show your work, then I had to grade each one.


Hey @khawks wow! What a cool thought!

Currently we’re working on a way to have a final answer for a Show Your Work so an answer key can be set, but this is very awesome and futuristic! :sunglasses: