Abstract:Summary SSL (semi‐supervised learning) is widely used in machine learning, which leverages labeled and unlabeled data to improve model performance. SSL aims to optimize class mutual information, but noisy pseudo‐labels introduce false class information due to the scarcity of labels. Therefore, these algorithms often need significant training time to refine pseudo‐labels for performance improvement iteratively. To tackle this challenge, we propose a novel plug‐and‐play method named Accelerating semi‐supervised learning via contrastive learning (ASCL). This method combines contrastive learning with uncertainty‐based selection for performance improvement and accelerates the convergence of SSL algorithms. Contrastive learning initially emphasizes the mutual information between samples as a means to decrease dependence on pseudo‐labels. Subsequently, it gradually turns to maximizing the mutual information between classes, aligning with the objective of semi‐supervised learning. Uncertainty‐based selection provides a robust mechanism for acquiring pseudo‐labels. The combination of the contrastive learning module and the uncertainty‐based selection module forms a virtuous cycle to improve the performance of the proposed model. Extensive experiments demonstrate that ASCL outperforms state‐of‐the‐art methods in terms of both convergence efficiency and performance. In the experimental scenario where only one label is assigned per class in the CIFAR‐10 dataset, the application of ASCL to Pseudo‐label, UDA (unsupervised data augmentation for consistency training), and Fixmatch benefits substantial improvements in classification accuracy. Specifically, the results demonstrate notable improvements in respect of 16.32%, 6.9%, and 24.43% when compared to the original outcomes. Moreover, the required training time is reduced by almost 50%.

Combating Medical Label Noise via Robust Semi-supervised Contrastive Learning

Semi-Supervised Relational Contrastive Learning

Hierarchical Semi-supervised Contrastive Learning for Contamination-Resistant Anomaly Detection

SCNet: semi-supervised and contrastive learning against noisy labels with two selection strategy approach

Deep semi-supervised learning with contrastive learning and partial label propagation for image data

MixCL: Pixel label matters to contrastive learning

Suppressing label noise in medical image classification using mixup attention and self-supervised learning

Selective-Supervised Contrastive Learning with Noisy Labels

Semi-supervised Contrastive Learning with Similarity Co-calibration.

Class-Aware Contrastive Semi-Supervised Learning

Label-noise-tolerant medical image classification via self-attention and self-supervised learning

PCLMix: Weakly Supervised Medical Image Segmentation via Pixel-Level Contrastive Learning and Dynamic Mix Augmentation

ASCL: Accelerating semi‐supervised learning via contrastive learning

Self-Ensembling Contrastive Learning for Semi-Supervised Medical Image Segmentation

Multi-modal contrastive mutual learning and pseudo-label re-learning for semi-supervised medical image segmentation

Affinity-Graph-Guided Contractive Learning for Pretext-Free Medical Image Segmentation with Minimal Annotation

Inherent Consistent Learning for Accurate Semi-supervised Medical Image Segmentation

Generating and Weighting Semantically Consistent Sample Pairs for Ultrasound Contrastive Learning

Semantic-Enhanced Supervised Contrastive Learning.

Class-aware Patch Based Contrastive Learning for Medical Image Segmentation

ChiMera: Learning with noisy labels by contrasting mixed-up augmentations