Abstract:In the face of complex natural images, existing deep clustering algorithms fall significantly short in terms of clustering accuracy when compared to supervised classification methods, making them less practical. This paper introduces an image clustering algorithm based on self-supervised pretrained models and latent feature distribution optimization, substantially enhancing clustering performance. It is found that: (1) For complex natural images, we effectively enhance the discriminative power of latent features by leveraging self-supervised pretrained models and their fine-tuning, resulting in improved clustering performance. (2) In the latent feature space, by searching for k-nearest neighbor images for each training sample and shortening the distance between the training sample and its nearest neighbor, the discriminative power of latent features can be further enhanced, and clustering performance can be improved. (3) In the latent feature space, reducing the distance between sample features and the nearest predefined cluster centroids can optimize the distribution of latent features, therefore further improving clustering performance. Through experiments on multiple datasets, our approach outperforms the latest clustering algorithms and achieves state-of-the-art clustering results. When the number of categories in the datasets is small, such as CIFAR-10 and STL-10, and there are significant differences between categories, our clustering algorithm has similar accuracy to supervised methods without using pretrained models, slightly lower than supervised methods using pre-trained models. The code linked algorithm is <a class="link-external link-https" href="https://github.com/LihengHu/semi" rel="external noopener nofollow">this https URL</a>.

Deep Self-paced Active Learning for Image Clustering

Self-paced Deep Clustering with Learning Loss

Deep Subspace Image Clustering Network with Self-Expression and Self-Supervision

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

Mejigclu: more effective jigsaw clustering for unsupervised visual representation learning

Graph-Based Semi-Supervised Deep Image Clustering With Adaptive Adjacency Matrix

Deep Density-based Image Clustering

Deep Structure and Attention Aware Subspace Clustering

Deep Clustering with Diffused Sampling and Hardness-aware Self-distillation

Deep image clustering: A survey

Self-learning for Face Clustering

Active Clustering Ensemble With Self-Paced Learning

Deep clustering based on embedded auto-encoder

Deep subspace clustering to achieve jointly latent feature extraction and discriminative learning

Pseudo-supervised Deep Subspace Clustering

Hyperspectral Image Classification Promotion Using Clustering Inspired Active Learning

Deep Attention-guided Graph Clustering with Dual Self-supervision

Self-Supervised Self-Organizing Clustering Network: A Novel Unsupervised Representation Learning Method

Deep Self-Supervised Attributed Graph Clustering for Social Network Analysis

Unsupervised Active Learning Based on Hierarchical Graph-Theoretic Clustering

Image Clustering with External Guidance