AUTHOR=Li Chen , Bai Menghan , Zhang Lipei , Xiao Ke , Song Wei , Zeng Hui 

TITLE=ACLMHA and FML: A brain-inspired kinship verification framework

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 16 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2022.1093071

DOI=10.3389/fnins.2022.1093071

ISSN=1662-453X

ABSTRACT=As an extended research direction of face recognition, kinship verification based on face image is an interesting yet challenging task, which aims to determine whether two individuals are kin-related or not based on their facial images. Face image based kinship verification can benefit many applications in real life, including: missing children search, family photo classification, kinship information mining, family privacy protection, etc. Studies presented thus far provide evidence that face kinship verification still faces many challenges. Hence in this paper we propose a novel kinship verification architecture, the main contribution are as follow: To boost the deep model to capture various and abundant local features from different local face regions, we propose an attention center learning guided multi-head attention mechanism to supervise the learning of attention weights and make different attention heads notice the characteristics of different regions. To better measure the potential similarity of features among relatives, we propose to introduce the feature relationship comparison module to measure the relationship between features at a deeper level. To combat the misclassification caused by single feature center loss, we propose a family-level multi-center loss to ensure a more proper intra/inter-class distance measurement for kinship verification. Extensive experiments are conducted on the widely-used kinship verification dataset FIW. Compared with other state-of-art methods, encouraging results are obtained which verify the effectiveness of our proposed method.