Unsupervised domain adaptation with Joint Adversarial Variational AutoEncoder
作者：
Highlights：
• Presenting a VAE-based transfer algorithm for cross-domain image classification.
• Adapting the marginal and conditional distributions by the Wasserstein distance.
• Combining adversarial learning techniques to robust the performance.
• Designing a three-classifier pseudo label generation strategy.
• Proposing a unique combined distance to map samples from the same class closely.
摘要
•Presenting a VAE-based transfer algorithm for cross-domain image classification.•Adapting the marginal and conditional distributions by the Wasserstein distance.•Combining adversarial learning techniques to robust the performance.•Designing a three-classifier pseudo label generation strategy.•Proposing a unique combined distance to map samples from the same class closely.
论文关键词：Domain adaptation,Deep learning,Joint distribution adaptation,Adversarial learning,Variational autoencoder
论文评审过程：Received 19 December 2021, Revised 14 May 2022, Accepted 14 May 2022, Available online 23 May 2022, Version of Record 2 June 2022.
论文官网地址：https://doi.org/10.1016/j.knosys.2022.109065
