Data augmentation for face recognition

详细信息    查看全文

• 作者：Jiang-Jing Lv^a ; ^b ; ^{lvjiangjing@cigit.ac.cn}Author Vitae ; Xiao-Hu Shao^b ; ^{shaoxiaohu@cigit.ac.cn}Author Vitae ; Jia-Shui Huang^b ; ^{huangjiashui@cigit.ac.cn}Author Vitae ; Xiang-Dong Zhou^b ; ^{zhouxiangdong@cigit.ac.cn}Author Vitae ; Xi Zhou^b ; ^{zhouxi@cigit.ac.cn}Author Vitae
• 关键词：Face recognition ; Data augmentation ; Landmark perturbation ; Image synthesis ; 3D reconstruction
• 刊名：Neurocomputing
• 出版年：2017
• 出版时间：22 March 2017
• 年：2017
• 卷：230
• 期：Complete
• 页码：184-196
• 全文大小：2336 K
• 卷排序：230

文摘

We present five data augmentation methods specific to face images. Landmark perturbation method is able to generate different kinds of transformed face images automatically. Different hairstyles and glasses of face image can be automatically synthesized. Face images with different poses and illuminations can be generated according to 3D face model.