Domain Adaptation by Topology Regularization

@article{Weeks2021DomainAB,
  title={Domain Adaptation by Topology Regularization},
  author={Deborah Weeks and Samuel Rivera},
  journal={ArXiv},
  year={2021},
  volume={abs/2101.12102}
}
Deep learning has become the leading approach to assisted target recognition, with transfer learning reducing the burden of requiring thousands of labeled samples. These approaches typically find intermediate feature or pixel representations where the target data match a related labeled source dataset with respect to some divergence measurement. While techniques have considered the global data structure through means such as KL-divergence, maximum mean discrepancy, and adversarial losses… Expand

References

SHOWING 1-10 OF 40 REFERENCES
Flexible deep transfer learning by separate feature embeddings and manifold alignment
Domain-Adversarial Training of Neural Networks
Wasserstein Generative Adversarial Networks
Adversarial Discriminative Domain Adaptation
Unsupervised Domain Adaptation by Backpropagation
Transferable Representation Learning with Deep Adaptation Networks
A Topology Layer for Machine Learning
...
1
2
3
4
...