("vectorized" in Norwegian)
A collaborative effort between me and my girlfriend Sophia Nearhood
Rather than using image recognition we're planning on converting all of the training data images to SVG. Then giving Vektorisert all the coordinates and a label of what text the image contains. You will use it by supplying Vektorisert with a string of text and it will output an array of generated coordinates for you. Those could just be pasted into an SVG, but we might go further with it and actually make a little web interface that would actually show you the image right away.
CC-BY 4.0 SA
If you enjoyed this project please consider sending a dollar or two as a thank you for the work :)