issues
search
pkuCactus
/
BDCN
The code for the CVPR2019 paper Bi-Directional Cascade Network for Perceptual Edge Detection
MIT License
345
stars
75
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
request for pretrained model params/bdcn_40000.pth
#65
Shirleyy32
opened
1 month ago
0
How to integrate multiple ground truth annotations into one image?
#64
Jiaxiong-huang
opened
7 months ago
0
有个问题没有解决
#63
kehan-hue
opened
8 months ago
0
Dice loss
#62
yx-yyds
opened
9 months ago
1
output feature
#61
shiwanghaha
opened
12 months ago
0
训练时loss一直为0
#60
Wei0205
opened
1 year ago
0
如何训练自己的数据集
#59
ChenMing-Wang
opened
1 year ago
0
训练速度
#58
cannonli7
opened
1 year ago
1
关于Multicue数据集的增强方式
#57
zhoujh2020
opened
2 years ago
0
Not as crisp results as shown in paper
#56
Ton-lee
opened
2 years ago
1
The reproduction result is poor
#55
liuxingyu123
closed
2 years ago
1
Dataset
#54
wsldy
opened
2 years ago
1
训练轮数问题
#53
1701000429
opened
2 years ago
1
about the SEM output shape
#52
tswizzle141
opened
2 years ago
1
BDCN-resnet版本,您可以公布吗
#51
lian666-ch
opened
2 years ago
1
pretrained model dose not work
#50
Li-yachuan
opened
2 years ago
1
对比实验数据问题
#49
studyhu
opened
2 years ago
2
Can u provide bdry.txt for better compare with other works
#48
BruceYu-Bit
opened
2 years ago
0
关于损失形式的一个疑惑
#47
lartpang
closed
3 years ago
2
No such file or directory: 'params/bdcn_final.pth'
#46
wenya1994
opened
3 years ago
3
Cannot reproduce the reported F-score
#45
ForawardStar
opened
3 years ago
9
MULTICUE checkpoint
#44
xavysp
opened
3 years ago
1
There is a error:FileNotFoundError: [Errno 2] No such file or directory: 'path_to/bsds500/BSR/BSDS500/data/test_pair.lst'
#43
sonicmlj
opened
3 years ago
1
网络输出的第7,8,9,10维度输出纯灰白图像?(Output pure gray image in dimension 7,8,9,10 of network output?)
#42
better-space
closed
3 years ago
1
How to calculate ODS, OIS
#41
WCJ1998
opened
3 years ago
1
Multicue groundtruth file for edge detection result evaluation
#40
rrryan2016
closed
3 years ago
5
multi-scale input
#39
ForawardStar
opened
3 years ago
2
How to prepare test_pair.lst for test.py?
#38
za13
opened
4 years ago
1
dataset
#37
zw-worker
opened
4 years ago
1
Why batch_size =1 ?
#36
Dreamerswile
closed
4 years ago
2
Non-Maximum Suppression
#35
ForawadStar
opened
4 years ago
1
Cannot find the caffe model
#34
tristanclare94
closed
4 years ago
2
Git clone now points to this repo, not PyTorch.
#33
stephenwithav
closed
4 years ago
0
About the data augmentation of PASCAL Context dataset?
#32
yuzhegao
opened
4 years ago
5
Results not reproducible
#31
balajiselvaraj1601
closed
4 years ago
4
The trained module?
#30
caihongzhiyue01
opened
5 years ago
0
Some questions about large training loss
#29
lly2111101
opened
5 years ago
7
训练速度很慢
#28
DreamerDoReMi
opened
5 years ago
6
Python 2.7?
#27
YacobBY
closed
5 years ago
4
links to datasets
#26
molefeskymleya
closed
5 years ago
1
Implementation of supervision (Y s2d) or (Y d2s)
#25
yuzhegao
opened
5 years ago
0
About the mean_bgr [104.00699, 116.66877, 122.67892]
#24
ourana35
opened
5 years ago
1
Training and Resuming
#23
huberthomas
closed
5 years ago
2
path_to/bsds500/BSR/BSDS500/data/voc_valtest.txt not found.
#22
Jillian555
opened
5 years ago
9
path_to/bsds500/BSR/BSDS500/data/test_id.txt not found.
#21
Jillian555
closed
5 years ago
0
Training Memory Consumption
#20
huberthomas
closed
5 years ago
2
Why use different lr for each layer? Not converge problem.
#19
anyuzoey
opened
5 years ago
9
what's error?
#18
Ruilkyu
closed
5 years ago
2
result
#17
LoveSimons
closed
5 years ago
4
用自己生成的数据训练loss高
#16
tangsipeng
closed
5 years ago
2
Next