SPICE: Semantic Pseudo-labeling for Image Clustering

Abstract

The similarity among samples and the discrepancy between clusters are two crucial aspects of image clustering. However, current deep clustering methods suffer from the inaccurate estimation of either feature similarity or semantic discrepancy. In this paper, we present a Semantic Pseudo-labeling-based Image ClustEring (SPICE) framework, which divides the clustering network into a feature model for measuring the instance-level similarity and a clustering head for identifying the cluster-level discrepancy. We design two semantics-aware pseudo-labeling algorithms, prototype pseudo-labeling, and reliable pseudo-labeling, which enable accurate and reliable self-supervision over clustering. Without using any ground-truth label, we optimize the clustering network in three stages: 1) train the feature model through contrastive learning to measure the instance similarity, 2) train the clustering head with the prototype pseudo-labeling algorithm to identify cluster semantics, and 3) jointly train the feature model and clustering head with the reliable pseudo-labeling algorithm to improve the clustering performance. Extensive experimental results demonstrate that SPICE achieves significant improvements (~10%) over existing methods and establishes the new state-of-the-art clustering results on six image benchmark datasets in terms of three popular metrics. Importantly, SPICE significantly reduces the gap between unsupervised and fully-supervised classification; e.g., there is only a 2% (91.8% vs 93.8%) accuracy difference on CIFAR-10. Our code has been made publically available at https://github.com/niuchuangnn/SPICE.

References in corpus (11)

• Distilling the Knowledge in a Neural Network
• Bootstrap your own latent: A new approach to self-supervised Learning
• Improved Regularization of Convolutional Neural Networks with Cutout
• Temporal Ensembling for Semi-Supervised Learning
• DivideMix: Learning with Noisy Labels as Semi-supervised Learning
• Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoders
• Pseudo-supervised Deep Subspace Clustering
• Deep Discriminative Clustering Analysis
• Clustering-friendly Representation Learning via Instance Discrimination and Feature Decorrelation
• HOME: High-Order Mixed-Moment-based Embedding for Representation Learning
• Self-Supervised Representation Learning With MUlti-Segmental Informational Coding (MUSIC)

Cited by in corpus (13)

• Twin Contrastive Learning for Online Clustering
• Learning Representation for Clustering via Prototype Scattering and Positive Sampling
• Self-supervised Contrastive Attributed Graph Clustering
• Unsupervised Contrastive Learning based Transformer for Lung Nodule Detection
• Forget Less, Count Better: A Domain-Incremental Self-Distillation Learning Benchmark for Lifelong Crowd Counting
• Self-supervised Image Clustering from Multiple Incomplete Views via Constrastive Complementary Generation
• Data clustering: a fundamental method in data science and management
• Grid Jigsaw Representation with CLIP: A New Perspective on Image Clustering
• Deep Transfer Clustering of Radio Signals
• PET-TURTLE: Deep Unsupervised Support Vector Machines for Imbalanced Data Clusters
• Improving Image Clustering through Sample Ranking and Its Application to remote--sensing images
• You Never Cluster Alone
• Cohort Organized Learning: Clustering Through Agreement