Hey everyone,
I’m excited to introduce OCRCourse! By utilizing advanced OCR pretraining with grounding capabilities, we’re pushing the boundaries of parsing-free methods for text recognition. Our data will be open-sourced soon, and both the model and paper are on the way!
Stay tuned for updates! 
Here’s an example running on a mobile emulator. 
: https://github.com/haikyu-kutappa-phapradana-tat-yuddha