A Survey of Multi-Ethnic Face Feature Recognition

Guo, Yijun; Ubul, Guljamal; Yadikar, Nurbiya; Mamut, Mutallip; Ubul, Kurban

doi:10.1145/3436369.3437412

Cited by 3 publications

(1 citation statement)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our experiments, all of the participants were Asian. Although different races result in differences in iris texture features [15], we envision that these differences have trivial impacts on the effectiveness of IriTrack. This is because the Daugman's integro-differential operator in IriTrack locates iris by detecting the path with the maximum change of pixel values, which is caused by abrupt change of pixel values of iris edge rather than iris textures.…”

Section: Discussionmentioning

confidence: 99%

IriTrack

Shen

Wei

Liao

et al. 2021

Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.

View full text Add to dashboard Cite

With a growing adoption of face authentication systems in various application scenarios, face Presentation Attack Detection (PAD) has become of great importance to withstand artefacts. Existing methods of face PAD generally focus on designing intelligent classifiers or customized hardware to differentiate between the image or video samples of a real legitimate user and the imitated ones. Although effective, they can be resource-consuming and suffer from performance degradation due to environmental changes. In this paper, we propose IriTrack, which is a simple and efficient PAD system that takes iris movement as a significant evidence to identify face artefacts. More concretely, users are required to move their eyes along with a randomly generated poly-line, where the resulting trajectories of their irises are used as an evidence for PAD i.e., a presentation attack will be identified if the deviation of one's actual iris trajectory from the given poly-line exceeds a threshold. The threshold is carefully selected to balance the latency and accuracy of PAD. We have implemented a prototype and conducted extensive experiments to evaluate the performance of the proposed system. The results show that IriTrack can defend against artefacts with moderate time and memory overheads.

show abstract

Section: Discussionmentioning

confidence: 99%