Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Improving Unsupervised Image Clustering With Robust Learning

About

Unsupervised image clustering methods often introduce alternative objectives to indirectly train the model and are subject to faulty predictions and overconfident results. To overcome these challenges, the current research proposes an innovative model RUC that is inspired by robust learning. RUC's novelty is at utilizing pseudo-labels of existing image clustering models as a noisy dataset that may include misclassified samples. Its retraining process can revise misaligned knowledge and alleviate the overconfidence problem in predictions. The model's flexible structure makes it possible to be used as an add-on module to other clustering methods and helps them achieve better performance on multiple datasets. Extensive experiments show that the proposed model can adjust the model confidence with better calibration and gain additional robustness against adversarial noise.

Sungwon Park, Sungwon Han, Sundong Kim, Danu Kim, Sungkyu Park, Seunghoon Hong, Meeyoung Cha• 2020

Related benchmarks

TaskDatasetResultRank
Image ClassificationSTL-10 (test)
Accuracy86.7
357
Image ClusteringCIFAR-10--
243
Image ClusteringSTL-10
ACC86.8
229
ClusteringCIFAR-10 (test)
Accuracy90
184
ClusteringSTL-10 (test)
Accuracy87
146
ClusteringCIFAR-100 (test)
ACC53
110
ClusteringCIFAR100 20
ACC0.533
93
Image ClassificationCIFAR-20
Mean Accuracy54.3
27
Unsupervised Image ClusteringCIFAR-20
Accuracy (Best)54.5
20
Unsupervised Image ClusteringImagenet 50
Best Accuracy78.5
2
Showing 10 of 10 rows

Other info

Code

Follow for update