arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2206.00893cs.CVcs.LG
收藏

利用二维变换的系统性知识

Leveraging Systematic Knowledge of 2D Transformations

University of Technology Sydney(悉尼科技大学);School of Electrical and Data Engineering(电气与数据工程学院)

Jiachen Kang, Wenjing Jia, Xiangjian He

AI总结:

针对深度学习模型在分布外场景下性能下降的问题,提出基于因果框架的合成数据训练方法获取二维变换系统性知识,并设计CED架构模拟人类视觉感知,显著提升协变量偏移下的分类准确率。

摘要:

现有的深度学习模型在计算机视觉任务中存在分布外(o.o.d.)性能下降的问题。相比之下,得益于所获知识的系统性,人类具有解释图像的卓越能力,即使图像中的场景很罕见。本工作专注于1)二维变换系统性知识的获取,以及2)能够在o.o.d.设置的图像分类任务中利用所学知识的架构组件。基于在因果框架下构建的合成数据集的新训练方法,深度神经网络从语义不同的领域(例如甚至从噪声中)获取知识,并在参数估计实验中表现出一定程度的系统性。基于此,设计了一种由分类器、估计器和标识符组成的新颖架构(缩写为“CED”)。通过模拟人类视觉感知中的“假设-验证”过程,CED在协变量偏移下的测试集上显著提高了分类准确率。

英文摘要:

The existing deep learning models suffer from out-of-distribution (o.o.d.) performance drop in computer vision tasks. In comparison, humans have a remarkable ability to interpret images, even if the scenes in the images are rare, thanks to the systematicity of acquired knowledge. This work focuses on 1) the acquisition of systematic knowledge of 2D transformations, and 2) architectural components that can leverage the learned knowledge in image classification tasks in an o.o.d. setting. With a new training methodology based on synthetic datasets that are constructed under the causal framework, the deep neural networks acquire knowledge from semantically different domains (e.g. even from noise), and exhibit certain level of systematicity in parameter estimation experiments. Based on this, a novel architecture is devised consisting of a classifier, an estimator and an identifier (abbreviated as "CED"). By emulating the "hypothesis-verification" process in human visual perception, CED improves the classification accuracy significantly on test sets under covariate shift.

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。

↑