Abstract:Unsupervised image clustering (UIC) is regularly employed to group images without manual annotation. One significant problem that occurs in the UIC context is that the visual-feature similarity across different semantic classes tends to introduce instance-dependent errors to clustering. The most successful recent approaches aimed at resolving this problem have focused on semisupervised reclustering, which utilizes reliable samples selected from the existing clusters. Despite this, virtually no previous work has considered the spatial consistency of the instance- and class-level representations which is crucial for error disambiguation. This makes it difficult to assess whether the selected reliable sample is reasonable. Accordingly, we propose a spatial consistency-based clustering (SCC) method to retain the alignment of representations learned from the instance and class levels in order to effectively select reliable samples from pre-existing clusters with errors thus leading to better clustering performance. More specifically, we first learn instance- and class-level representations by encouraging the semantic invariants of different instance augmentations and enforcing class alignment across semantically similar instances, respectively. We then assign instances with similar class-level representations to the same cluster to obtain the preliminary clusters. Subsequently we assess sample reliability by utilizing the spatial consistency constraint, which diffuses the class-level representations within the instance-level representation space. Finally, we employ semisupervised baselines combined with refinement techniques to perform reclustering based on the selected reliable samples. Extensive experimental results demonstrate that SCC can effectively obtain credible samples and outperform current SOTA clustering methods on the CIFAR-10 and CIFAR-100-20 benchmarks. The relevant code is available at https://github.com/RyanZhaoIc/SCC.git. (C)2022 Elsevier B.V. All rights reserved.

Combining core points and cluster-level semantic similarity for self-supervised clustering

Point Contrastive Prediction with Semantic Clustering for Self-Supervised Learning on Point Cloud Videos

A self-supervised learning of semantic feature consistency for image clustering

Effective Sample Pairs Based Contrastive Learning for Clustering

Learning the Relation between Similarity Loss and Clustering Loss in Self-Supervised Learning

Mejigclu: more effective jigsaw clustering for unsupervised visual representation learning

A Novel Image Clustering Algorithm Based on Supported Nearest Neighbors

Deep Contrastive Multi-view Clustering under Semantic Feature Guidance

Unsupervised node clustering via contrastive hard sampling

Unsupervised Contrastive Clustering Via Density Cluster Representative Combination

Improving unsupervised image clustering with spatial consistency

An End-to-End, Multi-Branch, Feature Fusion-Comparison Deep Clustering Method

Clustering-Aware Negative Sampling for Unsupervised Sentence Representation

Semantic Positive Pairs for Enhancing Visual Representation Learning of Instance Discrimination methods

Image Clustering Algorithm Based on Self-Supervised Pretrained Models and Latent Feature Distribution Optimization

Deep Clustering by Semantic Contrastive Learning

Improved Deep Clustering Model Based on Semantic Consistency for Image Clustering

Deep Image Clustering by Fusing Contrastive Learning and Neighbor Relation Mining.

Self-Supervised Clustering Models Based on BYOL Network Structure

Pairwise-Pixel Self-Supervised and Superpixel-Guided Prototype Contrastive Loss for Weakly Supervised Semantic Segmentation

Semantic-Enhanced Image Clustering