Controller-Guided Partial Label Consistency Regularization with Unlabeled Data

TL;DR

This paper introduces a controller-guided consistency regularization approach for partial label learning that leverages unlabeled data to improve performance, especially when partial annotations are insufficient.

Contribution

It proposes a novel method using a controller to guide consistency regularization at label and representation levels, dynamically adjusting thresholds to handle class imbalance and enhance existing PLL methods.

Findings

Achieves better performance in practical scenarios.

Modules can enhance existing PLL methods.

Effectively utilizes unlabeled data for improved learning.

Abstract

Partial label learning (PLL) learns from training examples each associated with multiple candidate labels, among which only one is valid. In recent years, benefiting from the strong capability of dealing with ambiguous supervision and the impetus of modern data augmentation methods, consistency regularization-based PLL methods have achieved a series of successes and become mainstream. However, as the partial annotation becomes insufficient, their performances drop significantly. In this paper, we leverage easily accessible unlabeled examples to facilitate the partial label consistency regularization. In addition to a partial supervised loss, our method performs a controller-guided consistency regularization at both the label-level and representation-level with the help of unlabeled data. To minimize the disadvantages of insufficient capabilities of the initial supervised model, we use the controller to estimate the confidence of each current prediction to guide the subsequent consistency regularization. Furthermore, we dynamically adjust the confidence thresholds so that the number of samples of each class participating in consistency regularization remains roughly equal to alleviate the problem of class-imbalance. Experiments show that our method achieves satisfactory performances in more practical situations, and its modules can be applied to existing PLL methods to enhance their capabilities.