Fully convolution networks are neural network composed exclusively of convolution and pooling layers. FCNs can be efficiently process images of any width and height. They are most useful for object detection and sematic segmentation because they only need to look at the images once.TrueFalse