alexgkendall / caffe-segnet

Implementation of SegNet: A Deep Convolutional Encoder-Decoder Architecture for Semantic Pixel-Wise Labelling

Home Page:http://mi.eng.cam.ac.uk/projects/segnet/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

The usage of upsample

408550969 opened this issue · comments

There is such a paragraph in segnet_sun.prototxt:
layer {
name: "upsample5"
type: "Upsample"
bottom: "pool5"
bottom: "pool5_mask"
top: "pool5_D"
upsample_param {
scale: 2
upsample_h: 23
upsample_w: 30
}
}
Why there are 2 bottom?When I change to one bottom,it will report an error:
Check failed: ExactNumBottomBlobs() == bottom.size() (2 vs. 1) Upsample Layer takes 2 bottom blob(s) as input.
Why must have 2 bottom? I just want to enlarge the source image.

int ExactNumBottomBlobs() const { return 2; } ;
one for feature maps, and one for the max-pooling indeices;

THANKS