Is it possible for A.I to redraw an image in a new context?

More specifically, suppose I take a photo of a notebook. I get the angle, lighting and perspective perfect. Now I copy an image I found online that contains text or handwriting. Can the A.I now draw this writing on my notebook?

Just asking if it’s possible, before I attempt to hire someone.

  • $\begingroup$ What do you mean by "draw this writing on my notebook"? Do you mean a computer physically drawing/painting an image onto a piece of paper? $\endgroup$ – Recessive Oct 7 '19 at 4:09
  • $\begingroup$ Is your goal only to repoduce "flat" images of text/drawings/photos in situ in placeholder images with existing or blank pages? What is the scale of number of placeholder images (the perfect shots of notebooks) and number of content images (the text/handwriting)? $\endgroup$ – Neil Slater Oct 7 '19 at 6:58
  • $\begingroup$ I have attempted to tidy up and clarify your question. Feel free to reverse or change the edit, especially if you think I have changed the meaning. $\endgroup$ – Neil Slater Oct 7 '19 at 7:03

This is theoretically possible given that you have enough labeled data to train a slightly modified general adversarial network (GAN). The training data would have to contain:

  • labeled pictures of notebooks
  • labeled pictures of text
  • labeled pictures of the notebook with text

The GAN will take the pictures of the notebooks and text as input, and will output a picture that will be used to calculate loss alongside the actual image of the notebook with text.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.