I\'m planning to use https://github.com/emedvedev/attention-ocr for implementing attention based OCR. I have images that I need to annotate, ie crop the text parts while also sa
