Looking for recommendations for text recognition from scans/photos → plain text with reasonably low error rate from my awful handwriting.
I prefer my notes digital to be searchable and harder to lose track of 😅 so I tend to type them directly or handwrite on a tablet, but I love writing with fountain pen and paper and I've always wanted to incorporate that better into my digital workflow.
Ideally open source, but will also consider paying for a service if good and doesn't sell all my data.
@petrichor I would very much be interested in the answer to this as well. Tesseract was something I used to convert a textbook to text file quite some time ago, but I recall it working okay but with some limitations: https://github.com/tesseract-ocr/tesseract It looks like some neural net capability has been added more recently.
I know Evernote does some OCR on images so that those become searchable. This comes close to what I would want a note taking solution to look like, except for the fact that my info is stored on Evernotes servers 😑
