Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / wvangansbeke/Unsupervised-Classification issues and pull requests

#140 - negative loss from SCANLoss

Issue - State: open - Opened by PeterDeSOM about 1 year ago - 1 comment

#139 - Minor bugfix

Pull Request - State: open - Opened by gihanjayatilaka about 1 year ago

#138 - Update env.yml

Pull Request - State: open - Opened by KKrishnAA12 about 1 year ago - 1 comment

#137 - Update env.yml

Pull Request - State: open - Opened by KKrishnAA12 about 1 year ago - 1 comment

#136 - Typo correction on cifar.py

Pull Request - State: closed - Opened by gihanjayatilaka over 1 year ago

#135 - Create timer

Pull Request - State: closed - Opened by pimber over 1 year ago

#132 - Help with installation

Issue - State: open - Opened by tungts1101 over 1 year ago - 1 comment

#131 - Solve the bad results caused by data imbalance?

Issue - State: closed - Opened by kkellyk over 1 year ago - 1 comment

#130 - Learned representations after pretext task

Issue - State: open - Opened by emessori almost 2 years ago - 2 comments

#129 - Normalization on custom dataset

Issue - State: closed - Opened by emessori almost 2 years ago - 1 comment

#127 - Accuracy becomes lower after self-labeling

Issue - State: closed - Opened by kkellyk almost 2 years ago

#126 - About "Pretext + Kmeans'

Issue - State: open - Opened by fi591 almost 2 years ago - 7 comments

#125 - Multi-GPU support

Issue - State: closed - Opened by doncat99 almost 2 years ago - 1 comment

#124 - The scan result is not ideal

Issue - State: open - Opened by kkellyk almost 2 years ago

#123 - Is it Unsupervised Learning or Semi-supervised Learning?

Issue - State: open - Opened by kkellyk almost 2 years ago

#122 - How to interpret confusion matrix in sacn?

Issue - State: closed - Opened by kkellyk almost 2 years ago

#121 - If an unlabeled dataset is used how should the target variable be handled?

Issue - State: closed - Opened by kkellyk almost 2 years ago - 2 comments

#120 - Is knn for validation?? (memory bank, target(label))

Issue - State: closed - Opened by edwardcho almost 2 years ago - 2 comments

#119 - Why do you need target (label) on fill_memory_bank??

Issue - State: closed - Opened by edwardcho almost 2 years ago - 1 comment

#118 - How to find labels as clustered results (ex: clustered images....)?

Issue - State: open - Opened by edwardcho almost 2 years ago - 3 comments

#117 - eval can't run the result

Issue - State: closed - Opened by kkellyk about 2 years ago - 1 comment

#116 - scan.py Error

Issue - State: closed - Opened by kkellyk about 2 years ago

#114 - Why is validation loss calculated without entropy weight?

Issue - State: closed - Opened by mazatov about 2 years ago - 2 comments

#113 - scan loss stuck, performs worse than pretext

Issue - State: open - Opened by mazatov about 2 years ago - 15 comments

#112 - Is there a simple way to train on unlabeled datas?

Issue - State: open - Opened by krNeko9t about 2 years ago

#111 - Was anyone able to train on their own dataset without labels?

Issue - State: open - Opened by Harsh9524 about 2 years ago - 3 comments

#110 - SelfLabel: Where is thresholding performed ?

Issue - State: closed - Opened by Yussef93 over 2 years ago

#109 - Extract visual features for new images

Issue - State: open - Opened by eleonoragualdoni over 2 years ago

#108 - About the hyper-parameters settting.

Issue - State: closed - Opened by blue-blue272 over 2 years ago - 3 comments

#107 - Regarding the consistency_loss in SCAN loss

Issue - State: closed - Opened by XiHuYan over 2 years ago - 2 comments

#106 - Can not import int_class, so I changed that part to int.

Pull Request - State: closed - Opened by masaishi over 2 years ago - 2 comments

#105 - HParams Overclustering

Issue - State: closed - Opened by Goschjann over 2 years ago

#104 - Clustering CIFAR-100

Issue - State: closed - Opened by avihu111 over 2 years ago - 2 comments

#103 - how to decide the number of clusters

Issue - State: closed - Opened by zhaotf16 over 2 years ago - 1 comment

#102 - how to evaluate the result of overclustering

Issue - State: closed - Opened by vmmm123 over 2 years ago - 2 comments

#101 - results after selflabel.py on CIFAR100 not include some classes?

Issue - State: closed - Opened by jimmy19991222 over 2 years ago - 2 comments

#100 - About Loss

Issue - State: closed - Opened by skr998 over 2 years ago - 4 comments

#99 - Results without false positives in the nearest neighbors

Issue - State: closed - Opened by skr998 over 2 years ago - 3 comments

#98 - Purpose of Memory Bank

Issue - State: closed - Opened by HenryPengZou over 2 years ago - 2 comments

#97 - Faiss assertion 'err == CUBLAS_STATUS_SUCCESS' failed

Issue - State: closed - Opened by MotiBaadror over 2 years ago - 1 comment

#96 - Motivation of semantic clustering through scan-loss

Issue - State: closed - Opened by wetliu almost 3 years ago - 2 comments

#95 - Dose label(targets) work while training?

Issue - State: closed - Opened by scteam1994 almost 3 years ago - 1 comment

#94 - Assertion error in SCAN process

Issue - State: closed - Opened by daminlee1 almost 3 years ago - 1 comment

#93 - Train on own dataset

Issue - State: open - Opened by fernandorovai almost 3 years ago - 27 comments

#92 - Overclustering with Majority Voting Details

Issue - State: closed - Opened by abrar360 almost 3 years ago - 1 comment

#91 - fine-tuning scan phase on custom dataset

Issue - State: closed - Opened by noamberg almost 3 years ago

#90 - Unable to unpack pretrained model tar files

Issue - State: closed - Opened by arnaudvl almost 3 years ago

#89 - config.yml for different datasets

Issue - State: closed - Opened by aakash-saboo almost 3 years ago - 2 comments

#87 - Download Links for CIFAR broken?

Issue - State: closed - Opened by ctataryn about 3 years ago - 3 comments

#86 - About the SCANLoss - Reopening Issue # 24

Issue - State: closed - Opened by rajat-tech-002 about 3 years ago - 2 comments

#84 - results on imagenet?

Issue - State: closed - Opened by seyeeet about 3 years ago - 2 comments

#84 - results on imagenet?

Issue - State: closed - Opened by seyeeet about 3 years ago - 2 comments

#83 - Settings for full ImageNet (not subset)

Issue - State: closed - Opened by mgwillia about 3 years ago - 5 comments

#83 - Settings for full ImageNet (not subset)

Issue - State: closed - Opened by mgwillia about 3 years ago - 5 comments

#82 - consistency loss and entropy loss

Issue - State: closed - Opened by ipheiman about 3 years ago - 1 comment

#81 - selflabel script is not performing well compare to scan script

Issue - State: closed - Opened by gauravjain10 about 3 years ago - 2 comments

#80 - CIFAR20 evaluation issue

Issue - State: closed - Opened by Xinyu-ustc about 3 years ago - 2 comments

#80 - CIFAR20 evaluation issue

Issue - State: closed - Opened by Xinyu-ustc about 3 years ago - 2 comments

#79 - Specifcy white background for dark mode enabled environments

Pull Request - State: closed - Opened by tlsJP about 3 years ago - 3 comments

#79 - Specifcy white background for dark mode enabled environments

Pull Request - State: closed - Opened by tlsJP about 3 years ago - 3 comments

#78 - Using pretrained for my own data

Issue - State: closed - Opened by dabasmoti about 3 years ago - 4 comments

#78 - Using pretrained for my own data

Issue - State: closed - Opened by dabasmoti about 3 years ago - 4 comments

#77 - About overclustering and underclustering

Issue - State: closed - Opened by michaelzfm about 3 years ago - 2 comments

#77 - About overclustering and underclustering

Issue - State: closed - Opened by michaelzfm about 3 years ago - 2 comments

#76 - Why not parallelize the implementation of simclr.py

Issue - State: closed - Opened by batradhruv about 3 years ago - 3 comments

#76 - Why not parallelize the implementation of simclr.py

Issue - State: closed - Opened by batradhruv about 3 years ago - 3 comments

#75 - Getting error while running scan.py at dataloader.

Issue - State: closed - Opened by sato-takahashi about 3 years ago - 3 comments

#75 - Getting error while running scan.py at dataloader.

Issue - State: closed - Opened by sato-takahashi about 3 years ago - 3 comments

#74 - Question on Backbone

Issue - State: closed - Opened by XingzhiZhou about 3 years ago - 2 comments

#73 - Corrected the `Slides` link in README.md

Pull Request - State: closed - Opened by ABD-01 about 3 years ago - 2 comments

#73 - Corrected the `Slides` link in README.md

Pull Request - State: closed - Opened by ABD-01 about 3 years ago - 2 comments

#72 - Getting error while running scan.py

Issue - State: closed - Opened by gauravjain10 about 3 years ago - 4 comments

#72 - Getting error while running scan.py

Issue - State: closed - Opened by gauravjain10 about 3 years ago - 4 comments

#71 - The question about reproducing the result of SCAN-loss based on the cifar-10.

Issue - State: closed - Opened by kiter-zero about 3 years ago - 2 comments

#71 - The question about reproducing the result of SCAN-loss based on the cifar-10.

Issue - State: closed - Opened by kiter-zero about 3 years ago - 2 comments

#70 - Overclustering

Issue - State: closed - Opened by abrar360 about 3 years ago - 1 comment

#69 - Repro Difficulties

Issue - State: closed - Opened by mgwillia about 3 years ago - 6 comments

#69 - Repro Difficulties

Issue - State: closed - Opened by mgwillia about 3 years ago - 6 comments

#68 - A couple of questions

Issue - State: closed - Opened by jyrr over 3 years ago - 3 comments

#68 - A couple of questions

Issue - State: closed - Opened by jyrr over 3 years ago - 3 comments

#67 - parallelization in simclr model

Issue - State: closed - Opened by hyungkwonko over 3 years ago - 2 comments

#67 - parallelization in simclr model

Issue - State: closed - Opened by hyungkwonko over 3 years ago - 2 comments

#66 - Number of classes much greater than the number of unique predictions obtained

Issue - State: closed - Opened by deshanadesai over 3 years ago - 6 comments

#66 - Number of classes much greater than the number of unique predictions obtained

Issue - State: closed - Opened by deshanadesai over 3 years ago - 6 comments

#65 - Train Scan with Imagenet backbone

Issue - State: closed - Opened by wtt0213 over 3 years ago - 1 comment

#64 - I want to train with my own dataset

Issue - State: closed - Opened by anewusername77 over 3 years ago - 8 comments

#64 - I want to train with my own dataset

Issue - State: closed - Opened by anewusername77 over 3 years ago - 8 comments

#63 - Questions about overclustering

Issue - State: closed - Opened by zhaoyun-ai over 3 years ago - 2 comments

#62 - About NeighborsDataset's neighbor and possible_neighbors

Issue - State: closed - Opened by woolpeeker over 3 years ago - 2 comments

#61 - pretext + kmeans results of ImageNet-50

Issue - State: open - Opened by ZhiyuanDang over 3 years ago - 9 comments

#60 - Unable to reproduce result from TUTORIAL.md

Issue - State: closed - Opened by wshinigamic over 3 years ago - 2 comments

#60 - Unable to reproduce result from TUTORIAL.md

Issue - State: closed - Opened by wshinigamic over 3 years ago - 2 comments

#59 - Question on SimCLR loss

Issue - State: closed - Opened by rajat-tech-002 over 3 years ago - 1 comment

#58 - Mnist evaluation

Issue - State: closed - Opened by eyalho over 3 years ago - 6 comments