Predicting Visual Exemplars of Unseen Classes for Zero-Shot Learning

被引:119
作者
Changpinyo, Soravit [1 ]
Chao, Wei-Lun [1 ]
Sha, Fei [1 ]
机构
[1] Univ Southern Calif, Los Angeles, CA 90007 USA
来源
2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV) | 2017年
基金
美国国家科学基金会;
关键词
DATABASE;
D O I
10.1109/ICCV.2017.376
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Leveraging class semantic descriptions and examples of known objects, zero-shot learning makes it possible to train a recognition model for an object class whose examples are not available. In this paper, we propose a novel zero-shot learning model that takes advantage of clustering structures in the semantic embedding space. The key idea is to impose the structural constraint that semantic representations must be predictive of the locations of their corresponding visual exemplars. To this end, this reduces to training multiple kernel-based regressors from semantic representation-exemplar pairs from labeled data of the seen object categories. Despite its simplicity, our approach significantly outperforms existing zero-shot learning methods on standard benchmark datasets, including the ImageNet dataset with more than 20,000 unseen categories.
引用
收藏
页码:3496 / 3505
页数:10
相关论文
共 51 条
[1]  
Akata Z., 2013, P IEEE C COMP VIS PA
[2]  
Akata Z., 2015, CVPR
[3]  
[Anonymous], 2011, ICML
[4]  
[Anonymous], 2016, IJCAI
[5]  
[Anonymous], 2014, CVPR
[6]  
[Anonymous], 2015, ICML
[7]  
[Anonymous], 2013, P NIPS
[8]  
[Anonymous], 2016, ECCV
[9]  
[Anonymous], 2014, PROC USENIX OSDI
[10]  
[Anonymous], 2015, ICCV