arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2105.04030cs.LG
收藏

更多贝叶斯:不确定性的域不变学习

A Bit More Bayesian: Domain-Invariant Learning with Uncertainty

University of Amsterdam(阿姆斯特丹大学);Inception Institute of Artificial Intelligence(Inception人工智能研究院)

Zehao Xiao, Jiayi Shen, Xiantong Zhen, Ling Shao, Cees G. M. Snoek

摘要:

由于领域迁移和目标领域数据的不可访问性所引起的不确定性,领域泛化具有挑战性。在本文中,我们解决这两个挑战的概率框架的基础上变分贝叶斯推理,通过将不确定性纳入神经网络的权重。将概率公式中的域不变性与变分贝叶斯推理相结合。这使我们能够以原则性的方式探索域不变学习。具体来说,我们推导出域不变的表示和分类器,这是共同建立在一个两层贝叶斯神经网络。我们实证证明了我们的建议的有效性,广泛使用的跨域视觉识别基准。消融研究验证了我们的贝叶斯治疗的协同效益时,共同学习领域不变的表示和分类器的领域泛化。此外,我们的方法在所有基准测试中始终提供最先进的平均精度。

英文摘要:

Domain generalization is challenging due to the domain shift and the uncertainty caused by the inaccessibility of target domain data. In this paper, we address both challenges with a probabilistic framework based on variational Bayesian inference, by incorporating uncertainty into neural network weights. We couple domain invariance in a probabilistic formula with the variational Bayesian inference. This enables us to explore domain-invariant learning in a principled way. Specifically, we derive domain-invariant representations and classifiers, which are jointly established in a two-layer Bayesian neural network. We empirically demonstrate the effectiveness of our proposal on four widely used cross-domain visual recognition benchmarks. Ablation studies validate the synergistic benefits of our Bayesian treatment when jointly learning domain-invariant representations and classifiers for domain generalization. Further, our method consistently delivers state-of-the-art mean accuracy on all benchmarks.

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。

↑