SeMi: When Imbalanced Semi-Supervised Learning Meets Mining Hard Examples

2025-01-10

AI Summary

Addressing semi-supervised learning under realistic class-imbalanced settings, this paper proposes the Entropy-Balanced Semi-Supervised Learning (EBSSL) framework. Methodologically: (1) It introduces hard-example mining—novel in this context—via a high-entropy identification strategy based on logits entropy differences; (2) it constructs a dynamically updated class-balanced memory bank, integrated with a confidence decay mechanism to enhance pseudo-label reliability; and (3) it jointly applies consistency regularization and pseudo-label optimization. The key contribution lies in systematically tackling both hard-example utilization and pseudo-label bias—two interdependent challenges in imbalanced semi-supervised learning. Extensive experiments demonstrate that EBSSL achieves significant improvements over state-of-the-art methods across multiple CISSL benchmarks; notably, under reverse-imbalanced scenarios, it outperforms the strongest baseline by up to 54.8% in absolute accuracy.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Semi-Supervised Learning (SSL) can leverage abundant unlabeled data to boost model performance. However, the class-imbalanced data distribution in real-world scenarios poses great challenges to SSL, resulting in performance degradation. Existing class-imbalanced semi-supervised learning (CISSL) methods mainly focus on rebalancing datasets but ignore the potential of using hard examples to enhance performance, making it difficult to fully harness the power of unlabeled data even with sophisticated algorithms. To address this issue, we propose a method that enhances the performance of Imbalanced Semi-Supervised Learning by Mining Hard Examples (SeMi). This method distinguishes the entropy differences among logits of hard and easy examples, thereby identifying hard examples and increasing the utility of unlabeled data, better addressing the imbalance problem in CISSL. In addition, we maintain a class-balanced memory bank with confidence decay for storing high-confidence embeddings to enhance the pseudo-labels' reliability. Although our method is simple, it is effective and seamlessly integrates with existing approaches. We perform comprehensive experiments on standard CISSL benchmarks and experimentally demonstrate that our proposed SeMi outperforms existing state-of-the-art methods on multiple benchmarks, especially in reversed scenarios, where our best result shows approximately a 54.8% improvement over the baseline methods.

Problem

• Semi-supervised Learning
• Class Imbalance
• Unlabeled Data Utilization

Innovation

• Semi-supervised Learning
• Imbalanced Data
• Memory Buffer