Pseudo-Labeling Curriculum for Unsupervised Domain Adaptation
This work addresses domain adaptation for computer vision tasks, offering an incremental improvement in handling noisy pseudo-labels.
The paper tackles the problem of false pseudo-labels degrading performance in unsupervised domain adaptation by proposing a curriculum based on density-based clustering to prioritize high-confidence samples early in training, achieving state-of-the-art results on benchmarks like Office-31, imageCLEF-DA, and Office-Home.
To learn target discriminative representations, using pseudo-labels is a simple yet effective approach for unsupervised domain adaptation. However, the existence of false pseudo-labels, which may have a detrimental influence on learning target representations, remains a major challenge. To overcome this issue, we propose a pseudo-labeling curriculum based on a density-based clustering algorithm. Since samples with high density values are more likely to have correct pseudo-labels, we leverage these subsets to train our target network at the early stage, and utilize data subsets with low density values at the later stage. We can progressively improve the capability of our network to generate pseudo-labels, and thus these target samples with pseudo-labels are effective for training our model. Moreover, we present a clustering constraint to enhance the discriminative power of the learned target features. Our approach achieves state-of-the-art performance on three benchmarks: Office-31, imageCLEF-DA, and Office-Home.