判别式
计算机科学
分类器(UML)
人工智能
成对比较
一致性(知识库)
特征(语言学)
模式识别(心理学)
域适应
机器学习
领域(数学分析)
集合(抽象数据类型)
光学(聚焦)
数学
哲学
数学分析
物理
程序设计语言
光学
语言学
作者
Zhongying Deng,Kaiyang Zhou,Yongxin Yang,Tao Xiang
出处
期刊:Cornell University - arXiv
日期:2021-11-06
标识
DOI:10.48550/arxiv.2111.03911
摘要
Most existing multi-source domain adaptation (MSDA) methods minimize the distance between multiple source-target domain pairs via feature distribution alignment, an approach borrowed from the single source setting. However, with diverse source domains, aligning pairwise feature distributions is challenging and could even be counter-productive for MSDA. In this paper, we introduce a novel approach: transferable attribute learning. The motivation is simple: although different domains can have drastically different visual appearances, they contain the same set of classes characterized by the same set of attributes; an MSDA model thus should focus on learning the most transferable attributes for the target domain. Adopting this approach, we propose a domain attention consistency network, dubbed DAC-Net. The key design is a feature channel attention module, which aims to identify transferable features (attributes). Importantly, the attention module is supervised by a consistency loss, which is imposed on the distributions of channel attention weights between source and target domains. Moreover, to facilitate discriminative feature learning on the target data, we combine pseudo-labeling with a class compactness loss to minimize the distance between the target features and the classifier's weight vectors. Extensive experiments on three MSDA benchmarks show that our DAC-Net achieves new state of the art performance on all of them.
科研通智能强力驱动
Strongly Powered by AbleSci AI