tfzhou ContrastiveSeg issues

tfzhou / ContrastiveSeg

ICCV2021 (Oral) - Exploring Cross-Image Pixel Contrast for Semantic Segmentation

MIT License

667 stars 88 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

CKPT

#69 geek-APTX4869 opened 6 months ago
0
Running the t code shows loss=nan when calculating the coco dataset.

#68 liaochuanlin opened 1 year ago
4
Dataloader distributed or not

#67 zhd2rng opened 1 year ago
0
@tfzhou So, the class 0 is the ignore index and not the background class? I mean, for the background class is applied also the contrastive loss. Right?

#66 zhangyuxuan321 opened 1 year ago
4
Code bug

#65 eraserNut opened 1 year ago
1
About the dimension of feature embedding

#64 chen-wolfberry opened 1 year ago
1
loss_contrast与loss_contrast_mem两个版本中的PixelContrastLoss区别

#63 IceHowe opened 1 year ago
0
problem with resuming training from checkpoint

#62 mailtohrishi opened 1 year ago
0
Example for max_views, max_samples?

#61 lennart-maack opened 2 years ago
2
code confusing me in def._sample_negative

#60 ck6698000 closed 2 years ago
4
Semi-Hard Example Sampling implementation

#59 nysp78 opened 2 years ago
3
How to reproduce the SoTA result of Cityscapes?

#58 JasiRose opened 2 years ago
1
Momentum Update

#57 Macc520 opened 2 years ago
2
how to reproduce the HRNet result trained on cityscapes trainval

#56 ICE-Bro opened 2 years ago
0
maybe bug in ./segmentor/trainer_contrastive.py

#55 eraserNut opened 2 years ago
5
loss_contrast

#54 Summer77723 opened 2 years ago
2
T-SNE of features visualization

#53 xiewende closed 2 years ago
1
The L2 normalization of features

#52 ccccly closed 2 years ago
2
Extremely slow validation during training + cpu overload

#51 davidceka opened 2 years ago
2
关于特征可视化的代码

#50 NJNUCS closed 2 years ago
1
Not able to run train file. Will please share the tips to run coco script step wise

#49 HamzaFarooq013 opened 2 years ago
0
about using your model in other dataset.

#48 Lemonweier closed 2 years ago
3
about the memory bank

#47 hhxzzy opened 2 years ago
0
How to migrate to my model？Looking forward to your answer, thanks.

#46 Pantherk1 opened 2 years ago
0
How to incorporate loss function into my model？

#45 202041600047 opened 2 years ago
3
about the decrease of the loss

#44 hhxzzy opened 2 years ago
8
About the hard anchor samping

#43 leeh43 opened 2 years ago
3
T-SNE of features visualization

#42 HuadongTang opened 2 years ago
7
an AttributeError occurred suddently during train with cityscapes dataset...

#41 shanhuhaifeng opened 2 years ago
7
can this code run on a lower torch version?

#40 hhxzzy closed 2 years ago
1
Dimension of feature embedding + full_res_stem

#39 oliverester closed 2 years ago
1
Can't find the code where it is saving loss weights during training

#38 faruknane closed 2 years ago
6
Cannot run the loss of loss_contrast_mem.py

#37 winycg opened 2 years ago
1
Questions about T-SNE visualization

#36 dongdongtong closed 2 years ago
3
why ignore zero?

#35 Junshan233 closed 2 years ago
3
About simple reproduction

#34 xianxuan-z closed 2 years ago
1
About args n_view,max_samples

#33 xianxuan-z opened 2 years ago
4
CUDA out of memory

#32 xiewende closed 2 years ago
1
Memory and semi-hard example mining code

#31 TheoPis opened 3 years ago
6
论文相似

#30 nlp-tokio closed 3 years ago
1
would you provide a command to run this code？Thanks a lot

#29 Brandy0k closed 2 years ago
1
network_stride & _dequeue_and_enqueue

#28 Darcy103 closed 2 years ago
2
Function _dequeue_and_enqueue() in train_contrastive.py

#27 Jiyang-Zheng closed 3 years ago
1
tSNE visualisation

#26 YuemingJin closed 2 years ago
1
Why ResNet-101 + DeepLab-V3 baseline performs so bad in your repo?

#25 MengzhangLI closed 3 years ago
1
Explanation of n_view

#24 HenryPengZou closed 2 years ago
3
when training code will be available?

#23 seyeeet closed 2 years ago
3
Question about hard sample mining

#22 curiosity654 opened 3 years ago
3
Explanation of n_view and Specific Dimension of the input of Contrastive function

#21 HenryPengZou closed 3 years ago
0
Question about the subtraction of max from inner product in the loss computation

#20 edwardpwtsoi closed 3 years ago
5