Hi, I'm really impressed with the text detection model, but I have a specific use case I'm interested in and I couldn't figure out if it's possible. I really want to be able to extract a mask representing the text pixels like in the image below. Is this possible to do with surya? I need something very accurate and very fine-detailed. Can surya output something like this? Thank you in advance.
