Human Saliency-Driven Patch-based Matching for Interpretable Post-mortem Iris Recognition

Aidan Boyd, Daniel Moreira, Andrey Kuehlkamp, Kevin Bowyer, Adam Czajka

Abstract:

Forensic iris recognition, as opposed to live iris recog-
nition, is an emerging research area that leverages the dis-
criminative power of iris biometrics to aid human examin-
ers in their efforts to identify deceased persons. As a ma-
chine learning-based technique in a predominantly human-
controlled task, forensic recognition serves as “back-up”
to human expertise in the task of post-mortem identifica-
tion. As such, the machine learning model must be (a)
interpretable, and (b) post-mortem-specific, to account for
changes in decaying eye tissue. In this work, we propose
a method that satisfies both requirements, and that ap-
proaches the creation of a post-mortem-specific feature ex-
tractor in a novel way employing human perception. We
first train a deep learning-based feature detector on post-
mortem iris images, using annotations of image regions
highlighted by humans as salient for their decision mak-
ing. In effect, the method learns interpretable features di-
rectly from humans, rather than purely data-driven features.
Second, regional iris codes (again, with human-driven
filtering kernels) are used to pair detected iris patches,
which are translated into pairwise, patch-based compari-
son scores. In this way, our method presents human ex-
aminers with human-understandable visual cues in order to
justify the identification decision and corresponding confi-
dence score. When tested on a dataset of post-mortem iris
images collected from 259 deceased subjects, the proposed
method places among the three best iris matchers, demon-
strating better results than the commercial (non-human-
interpretable) VeriEye approach. We propose a unique
post-mortem iris recognition method trained with human
saliency to give fully-interpretable comparison outcomes
for use in the context of forensic examination, achieving
state-of-the-art recognition performance.