issues
search
happynear
/
AMSoftmax
A simple yet effective loss function for face verification.
MIT License
485
stars
130
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
what is inner-product layer's output depended , How to calculate。
#35
ghost
closed
4 years ago
0
Alignment of the image in realtime
#34
robosina
closed
4 years ago
0
I put inner_product_layer and LabelSpecificAdd together,but it gets worse result.
#33
recordcode
opened
5 years ago
0
high similarity between abnormal face images
#32
jyshee
opened
5 years ago
2
Training without alignment
#31
borisgribkov
opened
5 years ago
0
flip in face_deploy_mirror_normalize.prototxt doesn't work properly
#30
robosina
opened
5 years ago
0
when set a mini-batch to model input the time of calculation won't decreade
#29
robosina
closed
5 years ago
0
Test Problem
#28
zys1994
opened
5 years ago
2
how to set the m and s when there are only 1000 identities?
#27
Chanbluky
closed
5 years ago
5
Where is fc6 in face_train_test.prototxt?
#26
Lin427
closed
5 years ago
2
how does the loss converge
#25
jyshee
closed
5 years ago
1
auto margin
#24
golunovas
closed
5 years ago
5
How to debug margin and scale?
#23
AITTSMD
opened
6 years ago
3
What is gnap prototxt for?
#22
twmht
closed
6 years ago
1
lfw准确率为99.73是如何得到的？使用的什么数据集！
#21
jianbaba
closed
6 years ago
12
Have you ever tried batch normalization?
#20
twmht
closed
6 years ago
4
Compare AMSoftmax with ArcFace
#19
basaltzhang
closed
6 years ago
1
CASIA-Webface dataset download link
#18
saurav4098
closed
6 years ago
31
iter_size问题
#17
wzj5133329
closed
5 years ago
1
The way of using AMSoftmax
#16
Anguliachao
closed
6 years ago
4
关于预训练模型在LFW数据集上准确率的问题
#15
xisi789
closed
6 years ago
12
CASIA-Webface dataset download link
#14
Yangel-hide
closed
6 years ago
9
draw_sphere
#13
xiaosuzhang
closed
6 years ago
3
using ./prototxt/face_solver.prototxt & ./prototxt/face_train_test.prototxt but not converge
#12
mingzhang96
closed
6 years ago
1
Can you send a verson of mnist dataset for amsoftmax, I modify a verson but when s=30 and m=0.4 it not converge and loss is also 16.811
#11
FlyEgle
closed
6 years ago
1
About pretrain model！
#10
zhangmo123
closed
6 years ago
1
face_train_test_wo_fn.prototxt
#9
dajidali010
closed
6 years ago
3
target logit and its curves
#8
xuezhongcailian
closed
6 years ago
5
关于LabelSpecificAdd问题
#7
jnulzl
closed
6 years ago
1
how to set the m when the feature without norm?
#6
Allan9977
closed
6 years ago
3
Setting of deploy
#5
banxia1994
closed
6 years ago
7
support caffe.binding?
#4
goxiaobu
closed
6 years ago
3
What‘s the performance of AMSoftmax on larger datasets？ MsCeleb， vggface2
#3
SeuTao
closed
6 years ago
5
the picture size is 96*96？？？
#2
taoyunuo
closed
6 years ago
1
about the weight norm
#1
zimenglan-sysu-512
closed
6 years ago
2