Domain Adaptation Using Pseudo Labels
This addresses domain adaptation for machine learning applications where labeled target data is unavailable, but it appears incremental as it builds on existing pseudo-label methods.
The paper tackles the problem of category misalignment in unsupervised domain adaptation by using a pretrained network with a multi-stage pseudo-label refinement procedure based on confidence, distance, and consistency, achieving results that demonstrate effectiveness compared to complex state-of-the-art techniques on multiple datasets.
In the absence of labeled target data, unsupervised domain adaptation approaches seek to align the marginal distributions of the source and target domains in order to train a classifier for the target. Unsupervised domain alignment procedures are category-agnostic and end up misaligning the categories. We address this problem by deploying a pretrained network to determine accurate labels for the target domain using a multi-stage pseudo-label refinement procedure. The filters are based on the confidence, distance (conformity), and consistency of the pseudo labels. Our results on multiple datasets demonstrate the effectiveness of our simple procedure in comparison with complex state-of-the-art techniques.