Journal of Shanghai Jiao Tong University (Science) ›› 2019, Vol. 24 ›› Issue (6): 699-705.doi: 10.1007/s12204-019-2132-0

• • 上一篇    下一篇

Transfer Learning Based on Joint Feature Matching and Adversarial Networks

ZHONG Haowen (钟昊文), WANG Chao (王超), TUO Hongya (庹红娅), HU Jian (胡健), QIAO Lingfeng (乔凌峰), JING Zhongliang (敬忠良)   

  1. (School of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai 200240, China)
  • 出版日期:2019-12-15 发布日期:2019-12-07
  • 通讯作者: TUO Hongya (庹红娅) E-mail: tuohy@sjtu.edu.cn

Transfer Learning Based on Joint Feature Matching and Adversarial Networks

ZHONG Haowen (钟昊文), WANG Chao (王超), TUO Hongya (庹红娅), HU Jian (胡健), QIAO Lingfeng (乔凌峰), JING Zhongliang (敬忠良)   

  1. (School of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai 200240, China)
  • Online:2019-12-15 Published:2019-12-07
  • Contact: TUO Hongya (庹红娅) E-mail: tuohy@sjtu.edu.cn

摘要: Domain adaptation and adversarial networks are two main approaches for transfer learning. Domain adaptation methods match the mean values of source and target domains, which requires a very large batch size during training. However, adversarial networks are usually unstable when training. In this paper, we propose a joint method of feature matching and adversarial networks to reduce domain discrepancy and mine domaininvariant features from the local and global aspects. At the same time, our method improves the stability of training. Moreover, the method is embedded into a unified convolutional neural network that can be easily optimized by gradient descent. Experimental results show that our joint method can yield the state-of-the-art results on three common public datasets.

关键词: transfer learning, adversarial networks, feature matching, domain-invariant features

Abstract: Domain adaptation and adversarial networks are two main approaches for transfer learning. Domain adaptation methods match the mean values of source and target domains, which requires a very large batch size during training. However, adversarial networks are usually unstable when training. In this paper, we propose a joint method of feature matching and adversarial networks to reduce domain discrepancy and mine domaininvariant features from the local and global aspects. At the same time, our method improves the stability of training. Moreover, the method is embedded into a unified convolutional neural network that can be easily optimized by gradient descent. Experimental results show that our joint method can yield the state-of-the-art results on three common public datasets.

Key words: transfer learning, adversarial networks, feature matching, domain-invariant features

中图分类号: