arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.08596cs.CV
收藏

面向领域泛化的特征风格化和领域感知对比学习

Feature Stylization and Domain-aware Contrastive Learning for Domain Generalization

Yonsei University(延世大学)

Seogkyu Jeon, Kibeom Hong, Pilhyeon Lee, Jewook Lee, Hyeran Byun

摘要:

领域泛化的目的是在不访问目标域的情况下增强模型对域移位的鲁棒性。由于可用于训练的源域有限,最近的方法集中于生成新域的样本。然而,它们要么在合成丰富域时与优化问题作斗争,要么导致类语义的扭曲。为此,我们提出了一个新的领域泛化框架,利用特征统计将原始特征风格化为具有新领域属性的特征。为了在风格化过程中保留类信息,我们首先将特征分解为高频和低频分量。之后,我们用从操纵统计量中采样的新域风格对低频分量进行风格化,同时保留高频分量中的形状线索。作为最后一步,我们重新合并两个组件,合成新的域功能。为了增强域鲁棒性,我们利用风格化的特征来保持模型在特征和输出方面的一致性。我们实现了与建议的域感知监督对比损失的特征一致性,这确保了域不变性,同时增加了类的可辨别性。实验结果表明,所提出的特征风格化和领域感知的对比度损失的有效性。通过定量比较,我们验证了我们的方法在两个基准,PACS和家庭的领先地位,现有的国家的最先进的方法。

英文摘要:

Domain generalization aims to enhance the model robustness against domain shift without accessing the target domain. Since the available source domains for training are limited, recent approaches focus on generating samples of novel domains. Nevertheless, they either struggle with the optimization problem when synthesizing abundant domains or cause the distortion of class semantics. To these ends, we propose a novel domain generalization framework where feature statistics are utilized for stylizing original features to ones with novel domain properties. To preserve class information during stylization, we first decompose features into high and low frequency components. Afterward, we stylize the low frequency components with the novel domain styles sampled from the manipulated statistics, while preserving the shape cues in high frequency ones. As the final step, we re-merge both components to synthesize novel domain features. To enhance domain robustness, we utilize the stylized features to maintain the model consistency in terms of features as well as outputs. We achieve the feature consistency with the proposed domain-aware supervised contrastive loss, which ensures domain invariance while increasing class discriminability. Experimental results demonstrate the effectiveness of the proposed feature stylization and the domain-aware contrastive loss. Through quantitative comparisons, we verify the lead of our method upon existing state-of-the-art methods on two benchmarks, PACS and Office-Home.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑