Lets say you want to detect something of a picture made with the bird perspective (ie with a drone). Can you create a dataset with the pictures and the self pre generated probability maps as labels, so that you train the input-image on the output-map?
$\begingroup$
$\endgroup$
2
-
$\begingroup$ What do you mean with "self pre generated probability maps as labels"? What is generating them? $\endgroup$– LeevoCommented Jun 24, 2019 at 15:14
-
$\begingroup$ For the dataset I need samples and labels, which are these maps. So I have to create them by myself, ie by writing a program on which I can click on the targets on the image and it returns such a „fake“ probability map $\endgroup$– T PiperCommented Jun 24, 2019 at 17:50
Add a comment
|