Add to Chrome
Add to Firefox
Add to EdgeHierarchical Variational Auto-Encoding for Unsupervised Domain Generalization
Feb 22, 2021
We address the task of domain generalization, where the goal is to train a predictive model based on a number of domains such that it is able to generalize to a new, previously unseen domain. We choose a generative approach within the framework of variational autoencoders and propose an unsupervised algorithm that is able to generalize to new domains without supervision. We show that our method is able to learn representations that disentangle domain-specific information from class-label specific information even in complex settings where an unobserved substructure is present in domains. Our interpretable method outperforms previously proposed generative algorithms for domain generalization and achieves competitive performance compared to state-of-the-art approaches, which are based on complex image-processing steps, on the standard domain generalization benchmark dataset PACS. Additionally, we proposed weak domain supervision which can further increase the performance of our algorithm in the PACS dataset.
