CroSel: Cross Selection of Confident Pseudo Labels for Partial-Label Learning
This work addresses label ambiguity in weakly supervised learning, offering a novel approach to improve model training by reducing noise in pseudo labels, though it appears incremental as it builds on existing identification-based methods.
The paper tackles the challenge of accurately identifying true labels in partial-label learning, where training examples have candidate label sets, by proposing CroSel, a method that uses cross selection and consistency regularization to select true labels with high precision, achieving over 90% accuracy and quantity on CIFAR-type datasets.
Partial-label learning (PLL) is an important weakly supervised learning problem, which allows each training example to have a candidate label set instead of a single ground-truth label. Identification-based methods have been widely explored to tackle label ambiguity issues in PLL, which regard the true label as a latent variable to be identified. However, identifying the true labels accurately and completely remains challenging, causing noise in pseudo labels during model training. In this paper, we propose a new method called CroSel, which leverages historical predictions from the model to identify true labels for most training examples. First, we introduce a cross selection strategy, which enables two deep models to select true labels of partially labeled data for each other. Besides, we propose a novel consistency regularization term called co-mix to avoid sample waste and tiny noise caused by false selection. In this way, CroSel can pick out the true labels of most examples with high precision. Extensive experiments demonstrate the superiority of CroSel, which consistently outperforms previous state-of-the-art methods on benchmark datasets. Additionally, our method achieves over 90\% accuracy and quantity for selecting true labels on CIFAR-type datasets under various settings.