MNIST数据库
计算机科学
判别式
人工智能
模式识别(心理学)
超球体
成对比较
特征(语言学)
子空间拓扑
机器学习
图像(数学)
嵌入
光学(聚焦)
一致性(知识库)
领域(数学分析)
任务(项目管理)
特征学习
班级(哲学)
样品(材料)
估计员
人工神经网络
数学
化学
管理
经济
哲学
数学分析
物理
光学
统计
色谱法
语言学
作者
Zan Gao,Yanbo Liu,Guangping Xu,Xianbin Wen
标识
DOI:10.1016/j.neucom.2020.06.147
摘要
In recent years, the domain adaption has received wide attention from machine learning communities because of differences in data distribution or the lack of training data in a practical machine learning task. In this work, we propose a P airwise A ttention N etwork ( PAN for short) for addressing cross-domain image recognition task. In this model, different local features and the global-feature are concatenated to obtain different attention estimators, and then they are combined to get the attention map. In this way, we can focus on the important parts of an image, and ignore the irrelative regions. Moreover, attention consistency is also embedded in PAN to make sure consistent interest regions in the same class. Besides, to improve the feature discrimination, an embedding discriminative subspace is learned where it maps positive sample pairs aligned in a hypersphere and negative sample pairs separated. Extensive experimental results on the MNIST-USPS, office, and Visda-2017 datasets demonstrate that PAN can outperform state-of-the-art methods in terms of average accuracy.
科研通智能强力驱动
Strongly Powered by AbleSci AI