Identity Document to Selfie Face Matching Across Adolescence

Matching live images (“selfies”) to images from ID documents is a problem that can arise in various applications. A challenging instance of the problem arises when the face image on the ID document is from early adolescence and the live image is from later adolescence. We explore this problem using a private dataset called Chilean Young Adult (CHIYA) dataset, where we match live face images taken at age 18-19 to face images on scanned ID documents created at ages 9 to 18. State-of-the-art deep learning face matchers (e.g., ArcFace) have relatively poor accuracy for document-to-selfie face matching. To achieve higher accuracy, we fine-tune the best available open-source model with triplet loss for a few-shot learning. Experiments show that our approach achieves higher accuracy than the DocFace+ model recently developed for this problem. Our fine-tuned model was able to improve the true acceptance rate for the most difficult (largest age span) subset from 62.92% to 96.67% at a false acceptance rate of 0.01%. Our fine-tuned model is available for use by other researchers.

1. Introduction

The problem of matching “live” or “selfie”¹

The term “selfie” has become used to describe a probe image in this context, even though the image may not be taken by the subject, meaning that it is not a literal selfie.

face images against images from an ID document (see Fig. 1) has been tackled in a number of previous works. Some previous works are from before the wave of deep learning algorithms in face recognition [3], [4], [24], [25]. There are also significant recent efforts involving deep learning algorithms [15], [21]. Figure 1:

Matching problem: selfie (left) and ID photo (right) in this case belong to the same person. In the selfie she is 18 years old, but in the ID photo she is 9 years old.