Jun-22-2018, 05:46 PM
(Jun-22-2018, 05:25 PM)nilamo Wrote: https://www.pyimagesearch.com/2018/06/18...-learning/ Facial recognition in python is fast enough to work while streaming video, without any noticeable delay. So if it's taking too long to do a static image, you might not be doing it the right way. Show some code, let's see what's going on.it was my previous question, i tried to delete all zeroes from image,and calculate it.
Yes machine learning can be fast, but im talking about pictures, they can be very very different. I need to calculate it not approximately, but precisely. It is good task,my suggestion was to delete all zeroes from left and right and remaining numbers to array, but it is too long*, even pathos multi cannot help
*my picture is 125 mb