Probability-Based Graph Embedding Cross-Domain and Class Discriminative Feature Learning for Domain Adaptation

Feature-based domain adaptation methods project samples from different domains into the same feature space and try to align the distribution of two domains to learn an effective transferable model. The vital problem is how to find a proper way to reduce the domain shift and improve the discriminability of features. To address the above issues, we propose a unified Probability-based Graph embedding Cross-domain and class Discriminative feature learning framework for unsupervised domain adaptation (PGCD). Specifically, we propose novel graph embedding structures to be the class discriminative transfer feature learning item and cross-domain alignment item, which can make the same-category samples compact in each domain, and fully align the local and global geometric structure across domains. Besides, two theoretical analyses are given to prove the interpretability of the proposed graph structures, which can further describe the relationships between samples to samples in single-domain and cross-domain transfer feature learning scenarios. Moreover, we adopt novel weight strategies via probability information to generate robust centroids in each proposed item to enhance the accuracy of transfer feature learning and reduce the error accumulation. Compared with the advanced approaches by comprehensive experiments, the promising performance on the benchmark datasets verify the effectiveness of the proposed model.

Automated summary

Feature-based domain adaptation methods aim to align the distribution of samples from different domains by projecting them into the same feature space, in order to learn a transferable model. The challenge lies in reducing the domain shift and improving the discriminability of features. To address these issues, we propose a unified Probability-based Graph embedding Cross-domain and class Discriminative feature learning framework (PGCD) for unsupervised domain adaptation. We introduce novel graph embedding structures as class discriminative transfer feature learning items and cross-domain alignment items, which compact same-category samples within each domain and align the local and global geometric structures across domains. The proposed model demonstrates promising performance on benchmark datasets compared to advanced approaches, validating its effectiveness.