arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1901.00282cs.CV
收藏

基于最小离散估计的深领域自适应算法

On Minimum Discrepancy Estimation for Deep Domain Adaptation

Queensland University of Technology (QUT)(昆士兰科技大学)

Mohammad Mahfujur Rahman, Clinton Fookes, Mahsa Baktashmotlagh, Sridha Sridharan

摘要:

在存在大量标记数据的情况下,深度学习(DL)在计算机视觉领域取得了非凡的胜利,特别是在对象分类和识别任务方面。然而,当训练图像和测试图像来自不同的分布或在训练图像和测试图像之间存在域移位时,DL并不总是表现良好。在没有标记的输入数据的情况下,它们也会受到影响。领域自适应(DA)方法已经被提出来弥补由于领域移位而导致的性能差。在本文中,我们提出了一种新的无监督的深领域自适应方法的基础上对齐的二阶统计量(协方差)以及最大均值差异的源和目标数据与双流卷积神经网络(CNN)。我们证明了所提出的方法在三个基准领域自适应数据集上实现图像分类的最新性能的能力:Office-31 [27],E-Home [37]和E-Caltech [8]。

英文摘要:

In the presence of large sets of labeled data, Deep Learning (DL) has accomplished extraordinary triumphs in the avenue of computer vision, particularly in object classification and recognition tasks. However, DL cannot always perform well when the training and testing images come from different distributions or in the presence of domain shift between training and testing images. They also suffer in the absence of labeled input data. Domain adaptation (DA) methods have been proposed to make up the poor performance due to domain shift. In this paper, we present a new unsupervised deep domain adaptation method based on the alignment of second order statistics (covariances) as well as maximum mean discrepancy of the source and target data with a two stream Convolutional Neural Network (CNN). We demonstrate the ability of the proposed approach to achieve state-of the-art performance for image classification on three benchmark domain adaptation datasets: Office-31 [27], Office-Home [37] and Office-Caltech [8].

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。

↑