4 ms·
Which ML training set data should I use to infer new applications from screenshots?
by tzakrajs 9y ago
Which ML training set data should I use to infer new applications from screenshots?
- dbcurtis 9y agohmmm... MNIST will get you digits... You jest, but actually, it might be an interesting problem to try to recreate program source from OCR'ed images. Using surrounding context, and then analyzing program flow, it might be possible to correct the inevitable transcription errors. The problem is just constrained enough to be interesting research.
- kingbirdy 9y agoOCR is good enough nowadays I cant see it having an issue with code screenshots - it's generally unstyled except for color, and in a fixed-width font.