Lighting-aware face frontalization for unconstrained face recognition
作者：
Highlights：
• Provide both lighting-recovered and lighting-normalized frontalized images.
• Basic frontalization with a generic 3D face model by the alignment of only five landmarks.
• Lighting recovered and normalized image filling by the symmetry of quotient image.
• LRFF method completes well with more sophisticated methods on the LFW benchmark.
• LNRR+LRA method outperforms the recent deep learning based methods on the MPIE database.
摘要
•Provide both lighting-recovered and lighting-normalized frontalized images.•Basic frontalization with a generic 3D face model by the alignment of only five landmarks.•Lighting recovered and normalized image filling by the symmetry of quotient image.•LRFF method completes well with more sophisticated methods on the LFW benchmark.•LNRR+LRA method outperforms the recent deep learning based methods on the MPIE database.
论文关键词：Face frontalization,Pose normalization,Illumination normalization,Unconstrained face recognition,Labeled face in the wild learning
论文评审过程：Received 4 December 2016, Revised 25 February 2017, Accepted 20 March 2017, Available online 22 March 2017, Version of Record 31 March 2017.
论文官网地址：https://doi.org/10.1016/j.patcog.2017.03.024
