issues
search
lorenmt
/
mtan
The implementation of "End-to-End Multi-Task Learning with Attention" [CVPR 2019].
https://shikun.io/projects/multi-task-attention-network
MIT License
665
stars
108
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Can you provide the code for pre-processing Cityscapes?
#63
HereNowL
closed
6 months ago
2
loss backward error
#62
Zhangxinyan666
closed
7 months ago
2
Used Mtan on depth estimation and object detection
#60
JiSuanJiDaWang
closed
2 years ago
5
Question about the structure of the encoder_block_att.
#59
JiSuanJiDaWang
closed
2 years ago
2
DWA的loss梯度爆炸问题
#58
E18301194
closed
2 years ago
2
Question about the pre-processed cityscapes
#57
heendung
closed
2 years ago
5
results inconsistency
#56
NoaGarnett
closed
2 years ago
19
关于Attention Module的设计
#55
puhan123
closed
2 years ago
4
why dwa didn't consider common scale of gradient?
#54
LMerCy
closed
2 years ago
8
Avg Cost accumulation
#53
sofiapvcp
closed
2 years ago
2
about visulization of the result
#52
ephisen
closed
2 years ago
1
about the uncertainty loss
#51
J-Wu97
closed
2 years ago
4
How do you pre-process NYUv2 dataset？
#50
tricomm
closed
2 years ago
7
The para Temperature of DWA
#49
ykeivn
closed
2 years ago
6
DWA
#48
long123524
closed
3 years ago
1
About the instantiation of the model
#47
LYT-C929
closed
3 years ago
1
Training MTAN-DeepLabv3
#46
Njuod
closed
3 years ago
10
Definition of multi-task loss function
#45
long123524
closed
3 years ago
5
Training with my own dataset
#44
Njuod
closed
3 years ago
16
Missing items in cityscapes/val/label_19
#43
zhanglijun95
closed
3 years ago
2
NYUv2 surface normals grey boundary
#42
NareshGuru77
closed
3 years ago
1
depth and segmentation label in cityscape
#41
TianhaoFu
closed
3 years ago
1
pre-processing CityScapes dataset google drive
#40
TianhaoFu
closed
3 years ago
11
Cannot unpickle dict_mean_std due to not mentioning python version
#39
janhenriklambrechts
closed
3 years ago
2
minimal NYU dataset
#38
MohsenFazaeli
closed
3 years ago
1
Avg cost function
#37
alexandruvesa
closed
3 years ago
2
Some problem of the implementation of task-specific attention networks.
#36
Watebear
closed
3 years ago
1
about uncert
#35
lelan-li
closed
3 years ago
0
why miou decrease when training?
#34
Manchery
closed
4 years ago
19
how to save and evaluate the model in the im2im_pred folder?
#33
syc10-09
closed
4 years ago
3
keras implementation
#32
tamaraalshekhli
closed
4 years ago
6
model_segnet_mtan prediction code
#31
ghost
closed
4 years ago
2
number of parameters
#30
cheer00
closed
4 years ago
2
Unstable evaluation results on NYUv2
#29
Manchery
closed
4 years ago
4
The Compute for Evaluation may not stable
#28
minygd
closed
4 years ago
6
Question about the architecture
#27
minygd
closed
4 years ago
2
about the backbone model
#26
liyangliu
closed
4 years ago
1
Cityscapes depth
#25
liyangliu
closed
4 years ago
13
Some questions about visual_decathlon
#24
chris20181220
closed
4 years ago
7
Mean of the sum with only one element
#23
joaopalotti
closed
4 years ago
1
My own dataset
#22
ntelo007
closed
4 years ago
1
Different Single-task results from the paper?
#21
xmasotto
closed
4 years ago
2
Conv init
#20
brdav
closed
4 years ago
2
NYU2_ dataset, npy file
#19
xingdw
closed
4 years ago
2
depth prediction result on cityscape dose not match with that reported in the paper
#18
ghost
closed
4 years ago
3
file structure
#17
xingdw
closed
4 years ago
4
If I want to train visual_decathlon without ImageNet,How to modify the code? Thanks!
#16
alwayslj
closed
4 years ago
1
cityscapes depth
#15
JingweiZhang12
closed
4 years ago
4
Results of directly running `model_segnet_single.py`
#14
lsongx
closed
4 years ago
4
Did you mask out pixels where the ground truth is missing?
#13
ghost
closed
4 years ago
2
Next