Saurabh Kumar
Nov 4 · 1 min read

Hey Karan,

Nice approach to use your own text detector rather than using bounding box provided by pytesseract. I also want to do something similar. My document dataset is quite diverse and I want to train text detection and classification of detected bounding box according to label. What are your thoughts on this?

Will it work on diverse document formats?

    Saurabh Kumar

    Written by