期刊
IEEE TRANSACTIONS ON IMAGE PROCESSING
卷 29, 期 -, 页码 6535-6548出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIP.2020.2991510
关键词
Semantics; Predictive models; Feature extraction; Training; Perturbation methods; Face recognition; Task analysis; Expression recognition; basic emotion; deep learning; autoencoder
资金
- Key Scientific Technological Innovation Research Project by the Ministry of Education, Zhejiang Provincial Natural Science Foundation of China [LR19F020004]
- National Natural Science Foundation of China [61751209]
- Zhejiang University K. P. Chao's High Technology Development Foundation
Different from many other attributes, facial expression can change in a continuous way, and therefore, a slight semantic change of input should also lead to the output fluctuation limited in a small scale. This consistency is important. However, current Facial Expression Recognition (FER) datasets may have the extreme imbalance problem, as well as the lack of data and the excessive amounts of noise, hindering this consistency and leading to a performance decreasing when testing. In this paper, we not only consider the prediction accuracy on sample points, but also take the neighborhood smoothness of them into consideration, focusing on the stability of the output with respect to slight semantic perturbations of the input. A novel method is proposed to formulate semantic perturbation and select unreliable samples during training, reducing the bad effect of them. Experiments show the effectiveness of the proposed method and state-of-the-art results are reported, getting closer to an upper limit than the state-of-the-art methods by a factor of 30% in AffectNet, the largest in-the-wild FER database by now.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据