Do We Really Need to Collect Millions of Faces for Effective Face Recognition?

被引:179
作者
Masi, Iacopo [1 ]
Anh Tuan Tran [1 ]
Hassner, Tal [2 ,3 ]
Leksut, Jatuporn Toy [1 ]
Medioni, Gerard [1 ]
机构
[1] USC, Inst Robot & Intelligent Syst, Los Angeles, CA USA
[2] USC, Inst Informat Sci, Los Angeles, CA 90071 USA
[3] Open Univ Israel, Raanana, Israel
来源
COMPUTER VISION - ECCV 2016, PT V | 2016年 / 9909卷
关键词
D O I
10.1007/978-3-319-46454-1_35
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Face recognition capabilities have recently made extraordinary leaps. Though this progress is at least partially due to ballooning training set sizes - huge numbers of face images downloaded and labeled for identity - it is not clear if the formidable task of collecting so many images is truly necessary. We propose a far more accessible means of increasing training data sizes for face recognition systems: Domain specific data augmentation. We describe novel methods of enriching an existing dataset with important facial appearance variations by manipulating the faces it contains. This synthesis is also used when matching query images represented by standard convolutional neural networks. The effect of training and testing with synthesized images is tested on the LFW and IJB-A (verification and identification) benchmarks and Janus CS2. The performances obtained by our approach match state of the art results reported by systems trained on millions of downloaded images.
引用
收藏
页码:579 / 596
页数:18
相关论文
共 47 条
[1]  
AbdAlmageed W, 2016, IEEE WINT CONF APPL
[2]  
[Anonymous], P INT C COMP VIS PAT
[3]  
[Anonymous], P INT C COMP VIS
[4]  
[Anonymous], P INT C COMP VIS PAT
[5]  
[Anonymous], P INT C COMP VIS PAT
[6]  
[Anonymous], P INT C COMP VIS PAT
[7]  
[Anonymous], WINT C APPL COMP VIS
[8]  
[Anonymous], EUROGRAPHICS 2014
[9]  
[Anonymous], P INT C COMP VIS PAT
[10]  
[Anonymous], P INT C COMP VIS PAT