Papers › Information Maximization Clustering via Multi-View Self-Labelling

Information Maximization Clustering via Multi-View Self-Labelling

12 Mar 2021arXiv:2103.07368archive 2025-07-28

Foivos Ntelemis, Yaochu Jin, Spencer A. Thomas

Image clustering is a particularly challenging computer vision task, which aims to generate annotations without human supervision. Recent advances focus on the use of self-supervised learning strategies in image clustering, by first learning valuable semantics and then clustering the image representations. These multiple-phase algorithms, however, increase the computational time and their final performance is reliant on the first stage. By extending the self-supervised approach, we propose a novel single-phase clustering method that simultaneously learns meaningful representations and assigns the corresponding annotations. This is achieved by integrating a discrete representation into the self-supervised paradigm through a classifier net. Specifically, the proposed clustering objective employs mutual information, and maximizes the dependency between the integrated discrete representation and a discrete probability distribution. The discrete probability distribution is derived though the self-supervised process by comparing the learnt latent representation with a set of trainable prototypes. To enhance the learning performance of the classifier, we jointly apply the mutual information across multi-crop views. Our empirical results show that the proposed framework outperforms state-of-the-art techniques with the average accuracy of 89.1% and 49.0%, respectively, on CIFAR-10 and CIFAR-100/20 datasets. Finally, the proposed method also demonstrates attractive robustness to parameter settings, making it ready to be applicable to other datasets.

PaperPDFCode

Code

foiv0s/imc-swav-pub officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

ClusteringImage ClassificationImage ClusteringSelf-Supervised Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Clustering CIFAR-10 IMC-SwAV (Best) ARI 0.8 #14 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Best) Accuracy 0.897 #14 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Best) Backbone ResNet-18 #14 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Best) NMI 0.818 #14 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Best) Train set Train #14 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Avg+-) ARI 0.79 #15 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Avg+-) Accuracy 0.891 #15 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Avg+-) Backbone ResNet-18 #15 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Avg+-) NMI 0.811 #15 of 40 Archive leaderboard report
Image Clustering CIFAR-10 IMC-SwAV (Avg+-) Train set Train #15 of 40 Archive leaderboard report
Image Clustering CIFAR-100 IMC-SwAV (Best) ARI 0.361 #12 of 30 Archive leaderboard report
Image Clustering CIFAR-100 IMC-SwAV (Best) Accuracy 0.519 #12 of 30 Archive leaderboard report
Image Clustering CIFAR-100 IMC-SwAV (Best) NMI 0.527 #12 of 30 Archive leaderboard report
Image Clustering CIFAR-100 IMC-SwAV (Best) Train Set Train #12 of 30 Archive leaderboard report
Image Clustering CIFAR-100 IMC-SwAV (Avg+-) ARI 0.337 #14 of 30 Archive leaderboard report
Image Clustering CIFAR-100 IMC-SwAV (Avg+-) Accuracy 0.49 #14 of 30 Archive leaderboard report
Image Clustering CIFAR-100 IMC-SwAV (Avg+-) NMI 0.503 #14 of 30 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Best) ARI 0.716 #10 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Best) Accuracy 0.853 #10 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Best) Backbone ResNet-18 #10 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Best) NMI 0.747 #10 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Best) Train Split Train #10 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Avg+-) ARI 0.685 #13 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Avg+-) Accuracy 0.831 #13 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Avg+-) Backbone ResNet-18 #13 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Avg+-) NMI 0.729 #13 of 29 Archive leaderboard report
Image Clustering STL-10 IMC-SwAV (Avg+-) Train Split Train #13 of 29 Archive leaderboard report
Image Clustering Tiny-ImageNet IMC-SwAV (Best) ARI 0.146 #4 of 14 Archive leaderboard report
Image Clustering Tiny-ImageNet IMC-SwAV (Best) Accuracy 0.282 #4 of 14 Archive leaderboard report
Image Clustering Tiny-ImageNet IMC-SwAV (Best) NMI 0.526 #4 of 14 Archive leaderboard report
Image Clustering Tiny-ImageNet IMC-SwAV (Avg+-) ARI 0.143 #5 of 14 Archive leaderboard report
Image Clustering Tiny-ImageNet IMC-SwAV (Avg+-) Accuracy 0.279 #5 of 14 Archive leaderboard report
Image Clustering Tiny-ImageNet IMC-SwAV (Avg+-) NMI 0.485 #5 of 14 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections