A typical domain adaptation approach is to adapt models trained on the annotated data in a source domain (e.g., sunny weather) for achieving high performance on the test data in a target domain (e.g., rainy weather). Whether the target contains a single homogeneous domain or multiple heterogeneous domains, existing works always assume that there exist clear distinctions between the domains, which is often not true in practice (e.g., changes in weather). We study an open compound domain adaptation (OCDA) problem, in which the target is a compound of multiple homogeneous domains without domain labels, reflecting realistic data collection from mixed and novel situations. We propose a new approach based on two technical insights into OCDA: 1) a curriculum domain adaptation strategy to bootstrap generalization across domains in a data-driven self-organizing fashion and 2) a memory module to increase the model's agility towards novel domains. Our experiments on digit classification, facial expression recognition, semantic segmentation, and reinforcement learning demonstrate the effectiveness of our approach.
@article{arxiv.1909.03403,
title = {Open Compound Domain Adaptation},
author = {Ziwei Liu and Zhongqi Miao and Xingang Pan and Xiaohang Zhan and Dahua Lin and Stella X. Yu and Boqing Gong},
journal= {arXiv preprint arXiv:1909.03403},
year = {2020}
}
Comments
To appear in CVPR 2020 as an oral presentation. Code, datasets and models are available at: https://liuziwei7.github.io/projects/CompoundDomain.html