Subclass heterogeneity aware loss for cross-spectral cross-resolution face recognition

被引：8
作者：
Ghosh S. [1 ]
Singh R. [2 ]
Vatsa M. [2 ]
机构：
[1] Department of Computer Science and Engineering, IIIT-Delhi, New Delhi
[2] Department of Computer Science and Engineering, IIT Jodhpur, Jodhpur
来源：
Vatsa, Mayank (mvatsa@iitj.ac.in) | 1600年 / Institute of Electrical and Electronics Engineers Inc.卷 / 02期
关键词：
Cross-spectral cross-resolution matching; Deep metric learning; Face recognition;
D O I：
10.1109/TBIOM.2020.2984324
中图分类号：
学科分类号：
摘要：
One of the most challenging scenarios of face recognition is matching images in presence of multiple covariates such as cross-spectrum and cross-resolution. In this paper, we propose a Subclass Heterogeneity Aware Loss (SHEAL) to train a deep convolutional neural network model such that it produces embeddings suitable for heterogeneous face recognition, both single and multiple heterogeneities. The performance of the proposed SHEAL function is evaluated on four databases in terms of the recognition performance as well as convergence in time and epochs. We observe that SHEAL not only yields state-of-the-art results for the most challenging case of Cross-Spectral Cross-Resolution face recognition, it also achieves excellent performance on homogeneous face recognition. © 2020 IEEE.
引用
收藏
页码：245 / 256
页数：11
相关论文
共 57 条
  • [1] Lu Z., Jiang X., Kot A., Deep coupled ResNet for low-resolution face recognition, IEEE Signal Process. Lett., 25, 4, pp. 526-530, (2018)
  • [2] Huo J., Gao Y., Shi Y., Yang W., Yin H., Heterogeneous face recognition by margin-based cross-modality metric learning, IEEE Trans. Comput., 48, 6, pp. 1814-1826, (2018)
  • [3] Wu X., He R., Sun Z., Tan T., A Light Cnn for Deep Face Representation with Noisy Labels, (2015)
  • [4] Parkhi O.M., Vedaldi A., Zisserman A., Deep face recognition, Proc. Brit. Mach. Vis. Conf. (BMVC), 1, pp. 6-19, (2015)
  • [5] O'Beirne S., What CCTV Users Need to Know about GDPR, (2017)
  • [6] Do Facial Recognition Cameras in Public Places Infringe on Our Privacy?, (2017)
  • [7] Zhu J.-Y., Zheng W.-S., Lai J.-H., Li S.Z., Matching NIR face to VIS face using transduction, IEEE Trans. Inf. Forensics Security, 9, 3, pp. 501-514, (2014)
  • [8] Nagpal S., Singh M., Singh R., Vatsa M., Noore A., Majumdar A., Face sketch matching via coupled deep transform learning, Proc. IEEE Int. Conf. Comput. Vis. (ICCV), pp. 5429-5438, (2017)
  • [9] Dhamecha T.I., Sharma P., Singh R., Vatsa M., On effectiveness of histogram of oriented gradient features for visible to near infrared face matching, Proc. IAPR ICPR, pp. 1788-1793, (2014)
  • [10] Grgic M., Delac K., Grgic S., SCface-Surveillance cameras face database, Multimedia Tools Appl., 51, 3, pp. 863-879, (2011)
123456