Pith. sign in

REVIEW 1 cited by

Aurora Guard: Real-Time Face Anti-Spoofing via Light Reflection

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1902.10311 v1 pith:TA6JJQGM submitted 2019-02-27 cs.CV

Aurora Guard: Real-Time Face Anti-Spoofing via Light Reflection

classification cs.CV
keywords lightmethodreflectionanti-spoofingaurorafacefurtherguard
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
abstract

In this paper, we propose a light reflection based face anti-spoofing method named Aurora Guard (AG), which is fast, simple yet effective that has already been deployed in real-world systems serving for millions of users. Specifically, our method first extracts the normal cues via light reflection analysis, and then uses an end-to-end trainable multi-task Convolutional Neural Network (CNN) to not only recover subjects' depth maps to assist liveness classification, but also provide the light CAPTCHA checking mechanism in the regression branch to further improve the system reliability. Moreover, we further collect a large-scale dataset containing $12,000$ live and spoofing samples, which covers abundant imaging qualities and Presentation Attack Instruments (PAI). Extensive experiments on both public and our datasets demonstrate the superiority of our proposed method over the state of the arts.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. SoK: Cybersecurity Assessment of Humanoid Ecosystem

    cs.CR 2025-08 conditional novelty 5.0

    A seven-layer humanoid security model with a 39x35 risk-weighted scoring method, demonstrated on Pepper, G1 EDU, and Digit.