Deep Least Squares Alignment for Unsupervised Domain Adaptation
arXiv:2111.02207
Abstract
Unsupervised domain adaptation leverages rich information from a labeled source domain to model an unlabeled target domain. Existing methods attempt to align the cross-domain distributions. However, the statistical representations of the alignment of the two domains are not well addressed. In this paper, we propose deep least squares alignment (DLSA) to estimate the distribution of the two domains in a latent space by parameterizing a linear model. We further develop marginal and conditional adaptation loss to reduce the domain discrepancy by minimizing the angle between fitting lines and intercept differences and further learning domain invariant features. Extensive experiments demonstrate that the proposed DLSA model is effective in aligning domain distributions and outperforms state-of-the-art methods.
BMVC 2021
References in corpus (5)
- Learning Transferable Features with Deep Adaptation Networks
- Deep Domain Confusion: Maximizing for Domain Invariance
- VisDA: The Visual Domain Adaptation Challenge
- Unsupervised Domain Adaptation via Discriminative Manifold Propagation
- Modified Distribution Alignment for Domain Adaptation with Pre-trained Inception ResNet