A domain adaptation framework based on black-box probing that blocks the risk of privacy leakage in practical applications. •. Distributionally adversarial ...

Source-free unsupervised domain adaptation is one class of practical deep learning methods which generalize in the target domain without transferring data ...

May 13, 2023 · Abstract:Source-free domain adaptation aims to adapt deep neural networks using only pre-trained source models and target data.

(c) Distributionally Adversarial Training: Fine-tune target model with unlabeled target data. Generate adversarial examples and retrain the target model. 3.

In our proposed method, we apply the covariate shift method to the training data expanded by the Daumé's feature augmentation method. In the experiment, we ...

Source-free and black-box domain adaptation via distributionally adversarial training ... Authors: Yucheng Shi; Kunhong Wu; Yahong Han; Yunfeng Shao; Bingshuai Li ...

A curated list of awesome source-free domain adaptation resources. Your contributions are always welcome! Contents. Shallow Methods; Image Classification ...

Congratulations! 06. June 2023. Yucheng Shi and Kunhong Wu's paper “Source-free and Black-box Domain Adaptation via Distributionally Adversarial Training” was ...

A novel domain adaptation method that fixes the classifier part of the model during adaptation and only fine-tune the remaining feature encoder part so that ...