Recognizing Families through Images with Pretrained Encoder

TL;DR

This paper explores kinship verification and retrieval using pretrained face encoders, achieving competitive results in a major challenge, and tests a generative model which did not improve performance.

Contribution

It evaluates multiple pretrained face encoders for kinship tasks and compares their effectiveness, including an attempt with StyleGAN2.

Findings

FaceNet and Siamese VGG-Face achieved top rankings in kinship verification and retrieval.

StyleGAN2 did not improve kinship recognition performance.

Pretrained face encoders are effective for kinship recognition tasks.

Abstract

Kinship verification and kinship retrieval are emerging tasks in computer vision. Kinship verification aims at determining whether two facial images are from related people or not, while kinship retrieval is the task of retrieving possible related facial images to a person from a gallery of images. They introduce unique challenges because of the hidden relations and features that carry inherent characteristics between the facial images. We employ 3 methods, FaceNet, Siamese VGG-Face, and a combination of FaceNet and VGG-Face models as feature extractors, to achieve the 9th standing for kinship verification and the 5th standing for kinship retrieval in the Recognizing Family in The Wild 2020 competition. We then further experimented using StyleGAN2 as another encoder, with no improvement in the result.