# LCGC: Learning from Consistency Gradient Conflicting for Class-Imbalanced Semi-Supervised Debiasing

arXiv: https://arxiv.org/abs/2504.06544 · HTML: https://mlsanity.com/paper/2504.06544/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 23 references that are in the map.
- First posted on arXiv 9 Apr 2025.

## Builds on (most cited first, 10 of 10)

- [FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence](https://mlsanity.com/paper/2001.07685/) (2020)
- [Class-Balanced Loss Based on Effective Number of Samples](https://mlsanity.com/paper/1901.05555/) (2019)
- [Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss](https://mlsanity.com/paper/1906.07413/) (2019)
- [Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning](https://mlsanity.com/paper/1704.03976/) (2017)
- [R-Drop: Regularized Dropout for Neural Networks](https://mlsanity.com/paper/2106.14448/) (2021)
- [CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learning](https://mlsanity.com/paper/2102.09559/) (2021)
- [Deep Stable Learning for Out-Of-Distribution Generalization](https://mlsanity.com/paper/2104.07876/) (2021)
- [ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shot](https://mlsanity.com/paper/2108.02385/) (2021)
- [ABC: Auxiliary Balanced Classifier for Class-imbalanced Semi-supervised Learning](https://mlsanity.com/paper/2110.10368/) (2021)
- [BaCon: Boosting Imbalanced Semi-supervised Learning via Balanced Feature-Level Contrastive Learning](https://mlsanity.com/paper/2403.12986/) (2024)

## Built on by (most cited first, 2 of 2)

- [SSS: Semi-Supervised SAM-2 with Efficient Prompting for Medical Imaging Segmentation](https://mlsanity.com/paper/2506.08949/) (2025)
- [Learning Dynamics of Logits Debiasing for Long-Tailed Semi-Supervised Learning](https://mlsanity.com/paper/2608.30699/) (2026)

## Abstract (arXiv)

Classifiers often learn to be biased corresponding to the class-imbalanced dataset, especially under the semi-supervised learning (SSL) set. While previous work tries to appropriately re-balance the classifiers by subtracting a class-irrelevant image's logit, but lacks a firm theoretical basis. We theoretically analyze why exploiting a baseline image can refine pseudo-labels and prove that the black image is the best choice. We also indicated that as the training process deepens, the pseudo-labels before and after refinement become closer. Based on this observation, we propose a debiasing scheme dubbed LCGC, which Learning from Consistency Gradient Conflicting, by encouraging biased class predictions during training. We intentionally update the pseudo-labels whose gradient conflicts with the debiased logits, representing the optimization direction offered by the over-imbalanced classifier predictions. Then, we debiased the predictions by subtracting the baseline image logits during testing. Extensive experiments demonstrate that LCGC can significantly improve the prediction accuracy of existing CISSL models on public benchmarks.