issues
search
wjgaas
/
DeepID2
use caffe to training a deepid2 net with softmax and contrastive loss, adding identity2verify_layer to translate a single sample into pairwise, thus mix the softmax and contrastive layer together
70
stars
30
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Check failed: ExactNumBottomBlobs() == bottom.size() (4 vs. 3) ContrastiveLoss Layer takes 4 bottom blob(s) as input.
#5
changgongcheng
opened
6 years ago
0
data prepare
#4
Dawson-huang
opened
6 years ago
0
Could you provide deploy prototxt?
#3
xhzcyc
opened
6 years ago
0
Where is your custom_layers.hpp
#2
veintiocho
opened
7 years ago
3
how to use deepid2
#1
anguoyang
opened
8 years ago
0