Correcting Noisy Multilabel Predictions: Modeling Label Noise through Latent Space Shifts

2025-02-20

AI Summary

To address label noise in multi-label classification arising from imperfect human annotations, this paper proposes a post-hoc correction framework that requires no model retraining. The method models multi-label noise as stochastic shifts in the latent space—a novel formulation—and constructs a unified unsupervised and semi-supervised correction mechanism grounded in deep generative modeling, jointly leveraging uncertainty estimation and latent variable inference. Designed to be compatible with existing denoising strategies, it significantly reduces computational overhead. Extensive experiments across diverse noise types and intensities demonstrate consistent superiority over state-of-the-art baselines, achieving new SOTA performance on multiple benchmark datasets. Ablation studies and sensitivity analyses rigorously validate the effectiveness of each component and confirm the overall robustness of the framework.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Reasoning under Uncertainty: Stochastic Optimization
• Computer Vision: Multi-modal Vision

Application Category

• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Noise in data appears to be inevitable in most real-world machine learning applications and would cause severe overfitting problems. Not only can data features contain noise, but labels are also prone to be noisy due to human input. In this paper, rather than noisy label learning in multiclass classifications, we instead focus on the less explored area of noisy label learning for multilabel classifications. Specifically, we investigate the post-correction of predictions generated from classifiers learned with noisy labels. The reasons are two-fold. Firstly, this approach can directly work with the trained models to save computational resources. Secondly, it could be applied on top of other noisy label correction techniques to achieve further improvements. To handle this problem, we appeal to deep generative approaches that are possible for uncertainty estimation. Our model posits that label noise arises from a stochastic shift in the latent variable, providing a more robust and beneficial means for noisy learning. We develop both unsupervised and semi-supervised learning methods for our model. The extensive empirical study presents solid evidence to that our approach is able to consistently improve the independent models and performs better than a number of existing methods across various noisy label settings. Moreover, a comprehensive empirical analysis of the proposed method is carried out to validate its robustness, including sensitivity analysis and an ablation study, among other elements.

Problem

• Correcting noisy multilabel predictions
• Modeling label noise through latent shifts
• Improving robustness in noisy label learning

Innovation

• Deep generative approaches
• Latent space shifts
• Unsupervised and semi-supervised methods