Multi-scale deep relational reasoning for facial kinship verification

作者：

Highlights：

• We propose a relation network for kinship identification that explores relationships of multi-scale features.

• We first use two shared parameter convolutional neural networks to extract different scale features from a pair of face images. These features will be used to provide context information and global information about face images

• Compared with most state-of-the-arts, experimental results demonstrate that our proposed methods obtain superior performances on two benchmark datasets.

摘要

•We propose a relation network for kinship identification that explores relationships of multi-scale features.•We first use two shared parameter convolutional neural networks to extract different scale features from a pair of face images. These features will be used to provide context information and global information about face images•Compared with most state-of-the-arts, experimental results demonstrate that our proposed methods obtain superior performances on two benchmark datasets.

论文关键词：Facial analysis,Kinship verification,Biometrics,Feature learning

论文评审过程：Received 9 November 2019, Revised 11 April 2020, Accepted 3 July 2020, Available online 6 July 2020, Version of Record 1 November 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107541