wvangansbeke Unsupervised-Classification issues

wvangansbeke / Unsupervised-Classification

SCAN: Learning to Classify Images without Labels, incl. SimCLR. [ECCV 2020]

https://arxiv.org/abs/2005.12320

Other

1.37k stars 268 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Is it normal that i got a loss which is negative with the entropy_weight=5.0?

#141 Frankie0403 opened 1 year ago
2
negative loss from SCANLoss

#140 PeterDeSOM opened 1 year ago
1
Minor bugfix

#139 gihanjayatilaka opened 1 year ago
0
Update env.yml

#138 KKrishnAA12 opened 1 year ago
1
Update env.yml

#137 KKrishnAA12 opened 1 year ago
1
Typo correction on cifar.py

#136 gihanjayatilaka closed 1 year ago
0
Create timer

#135 pimber closed 1 year ago
0
SimCLR loss and nearest neighbors computed on hidden features?

#134 emessori opened 1 year ago
0
Help with installation

#132 tungts1101 opened 1 year ago
1
Solve the bad results caused by data imbalance?

#131 kkellyk closed 1 year ago
1
Learned representations after pretext task

#130 emessori opened 1 year ago
2
Normalization on custom dataset

#129 emessori closed 1 year ago
1
Overcoming uncertainity after scan phase by dynamically lowering accuracy threshold for self-labeling

#128 TomasPlachy opened 2 years ago
0
Accuracy becomes lower after self-labeling

#127 kkellyk closed 2 years ago
0
About "Pretext + Kmeans'

#126 fi591 opened 2 years ago
7
Multi-GPU support

#125 doncat99 closed 1 year ago
1
The scan result is not ideal

#124 kkellyk opened 2 years ago
0
Is it Unsupervised Learning or Semi-supervised Learning?

#123 kkellyk opened 2 years ago
0
How to interpret confusion matrix in sacn?

#122 kkellyk closed 2 years ago
0
If an unlabeled dataset is used how should the target variable be handled?

#121 kkellyk closed 2 years ago
2
Is knn for validation?? (memory bank, target(label))

#120 edwardcho closed 2 years ago
2
Why do you need target (label) on fill_memory_bank??

#119 edwardcho closed 2 years ago
1
How to find labels as clustered results (ex: clustered images....)?

#118 edwardcho opened 2 years ago
3
eval can't run the result

#117 kkellyk closed 2 years ago
1
scan.py Error

#116 kkellyk closed 2 years ago
0
Do you provide the parameters of the RandAgument mentioned in the paper?

#115 kkellyk closed 2 years ago
0
Why is validation loss calculated without entropy weight?

#114 mazatov closed 2 years ago
2
scan loss stuck, performs worse than pretext

#113 mazatov opened 2 years ago
15
Is there a simple way to train on unlabeled datas?

#112 krNeko9t opened 2 years ago
0
Was anyone able to train on their own dataset without labels?

#111 Harsh9524 opened 2 years ago
3
SelfLabel: Where is thresholding performed ?

#110 Yussef93 closed 2 years ago
0
Extract visual features for new images

#109 eleonoragualdoni opened 2 years ago
0
About the hyper-parameters settting.

#108 blue-blue272 closed 2 years ago
3
Regarding the consistency_loss in SCAN loss

#107 XiHuYan closed 2 years ago
2
Can not import int_class, so I changed that part to int.

#106 masaishi closed 2 years ago
2
HParams Overclustering

#105 Goschjann closed 2 years ago
0
Clustering CIFAR-100

#104 avihu111 closed 2 years ago
2
how to decide the number of clusters

#103 zhaotf16 closed 2 years ago
1
how to evaluate the result of overclustering

#102 vmmm123 closed 2 years ago
2
results after selflabel.py on CIFAR100 not include some classes?

#101 jimmy19991222 closed 2 years ago
2
About Loss

#100 skr998 closed 2 years ago
4
Results without false positives in the nearest neighbors

#99 skr998 closed 2 years ago
3
Purpose of Memory Bank

#98 HenryPengZou closed 2 years ago
2
Faiss assertion 'err == CUBLAS_STATUS_SUCCESS' failed

#97 MotiBaadror closed 2 years ago
1
Motivation of semantic clustering through scan-loss

#96 wetliu closed 2 years ago
2
Dose label(targets) work while training?

#95 scteam1994 closed 3 years ago
1
Assertion error in SCAN process

#94 daminlee1 closed 2 years ago
1
Train on own dataset

#93 fernandorovai opened 3 years ago
27
Overclustering with Majority Voting Details

#92 abrar360 closed 2 years ago
1
fine-tuning scan phase on custom dataset

#91 noamberg closed 3 years ago
0