Dark Reading is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Comments
Researchers Show Vulnerabilities in Facial Recognition
Newest First  |  Oldest First  |  Threaded View
JohnW891
JohnW891,
User Rank: Apprentice
8/12/2019 | 2:04:18 PM
There's More To This
Face recognition is a part of face biometrics that includes face authentication, a materially different approach. Recognition is based entirely on image matching. There is no liveness detection, the primary differentiator that allows an authentication system to both verify what the camera sees with what is on file *and* determine whether that "correct" person is actually alive at the time of the access request. But, not in that order.

Liveness detection (liveness.com) is key to actual authentication because is assess, literally, dozens of unique human attributes to determine that *together* they add up to a real human. If the liveness test isn't passed first, there is no need to match images. This prevents things like detailed masks, 4K video, etc. from standing in for the real person.

In addition, the data acquisition method for recognition is based on 2D images. In true authentication, 3D images provide far more signal to work with. This is why Face ID, even though it has no liveness detection, works better most of the time than the legacy algorithms that have powered recognition for, literally, decades. It has more to work with to make a more accurate match. But, as was seen (just look it up on YouTube), it's easy to spoof. And Apple will *never* subject it's tech to a transparent test, so we only have social proof to work with.

This is old news.


Edge-DRsplash-10-edge-articles
I Smell a RAT! New Cybersecurity Threats for the Crypto Industry
David Trepp, Partner, IT Assurance with accounting and advisory firm BPM LLP,  7/9/2021
News
Attacks on Kaseya Servers Led to Ransomware in Less Than 2 Hours
Robert Lemos, Contributing Writer,  7/7/2021
Commentary
It's in the Game (but It Shouldn't Be)
Tal Memran, Cybersecurity Expert, CYE,  7/9/2021
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
Incorporating a Prevention Mindset into Threat Detection and Response
Threat detection and response systems, by definition, are reactive because they have to wait for damage to be done before finding the attack. With a prevention-mindset, security teams can proactively anticipate the attacker's next move, rather than reacting to specific threats or trying to detect the latest techniques in real-time. The report covers areas enterprises should focus on: What positive response looks like. Improving security hygiene. Combining preventive actions with red team efforts.
Flash Poll
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2022-31650
PUBLISHED: 2022-05-25
In SoX 14.4.2, there is a floating-point exception in lsx_aiffstartwrite in aiff.c in libsox.a.
CVE-2022-31651
PUBLISHED: 2022-05-25
In SoX 14.4.2, there is an assertion failure in rate_init in rate.c in libsox.a.
CVE-2022-29256
PUBLISHED: 2022-05-25
sharp is an application for Node.js image processing. Prior to version 0.30.5, there is a possible vulnerability in logic that is run only at `npm install` time when installing versions of `sharp` prior to the latest v0.30.5. If an attacker has the ability to set the value of the `PKG_CONFIG_PATH` e...
CVE-2022-26067
PUBLISHED: 2022-05-25
An information disclosure vulnerability exists in the OAS Engine SecureTransferFiles functionality of Open Automation Software OAS Platform V16.00.0112. A specially-crafted series of network requests can lead to arbitrary file read. An attacker can send a sequence of requests to trigger this vulnera...
CVE-2022-26077
PUBLISHED: 2022-05-25
A cleartext transmission of sensitive information vulnerability exists in the OAS Engine configuration communications functionality of Open Automation Software OAS Platform V16.00.0112. A targeted network sniffing attack can lead to a disclosure of sensitive information. An attacker can sniff networ...