Re:Finding Waldo Using Semantic Segmentation and Tiramisu
by p1esk on Sunday October 08, 2017 @08:04PM
(#15430708)
they can do calculations on a pixel-by-pixel basis, as opposed to standard methods which use are forced to compress information along the way through pooling/strided convolutions
I'm not sure I understand your point. Why do you think dilated filters produce a better global view than pooling?
I'm not sure I understand your point. Why do you think dilated filters produce a better global view than pooling?