查看原文
其他

你看,我的心没变,还能解锁!

LearnAndRecord 2022-07-26

心脏也能当密码刷?

人脸识别、指纹识别都弱爆了


指纹、瞳孔、面部特征、声纹等特征能用来做身份验证并不为奇,然而大部分人不知道的是,每个人心脏的形状和指纹、瞳孔一样,也是独一无二的。近日,美国布法罗大学(University at Buffalo)研究者开发了一套“心脏扫描认证系统”,号称能利用心脏形状、大小和跳动特征来鉴别人的身份,用于身份认证。

Like fingerprints and facial recognition, the shape and beat of your heart can be used to verify your identity.


Biometric Identifies You in a Heartbeat


60-SECOND SCIENCE

Smartphones can already verify your identity by scanning your fingerprint or your face. But there's something inside you that's also unique proof that you're you.


Researchers have developed a device that uses low-powered Doppler radar* to decode the unique geometry of a user's heart…and how it squeezes and swells as it beats. That biometric information can authenticate the identity of the person under scrutiny*.


利用一台特殊的多普勒雷达,“心脏扫描器”就能收集到人心脏的活动特征,并转换成数字信号传输给身份认证系统。


*Doppler radar 多普勒雷达


*scrutiny:the careful and detailed examination of something in order to get information about it 细看,仔细审查

The government's record will be subjected to/come under (close) scrutiny in the weeks before the election.

政府的所作所为将在选举前的几个星期接受审查。

And if the wrong person sits down: "It would recognize that there is a heart, but the geometry of my heart is different from yours so it would still log me out." Wenyao Xu is a computer scientist at the University at Buffalo who helped develop the tech.


He says this unusual biometric is robust, because it's actually two different biometrics. The first is the shape of your heart—a 'biologic' biometric or static trait, like fingerprints or iris* patterns. The second is the beating of your heart: what's called a 'behavioral' biometric—which analyzes a dynamic process, and can be harder to spoof.


*iris:the coloured circular part of that eye that surrounds the black pupil (= central part) 虹膜


So far they've tested the system on about 100 people—with an accuracy of over 98 percent. They'll present the results this month at the MobiCom conference in Utah. [Chen Song et al, Cardiac Scan: A Non-contact and Continuous Heart-based User Authentication System]


为了测试这台雷达装置,研究者找了78人进行测试,发现他们的平均扫描精准度能达到98.61%,误差率4.42%。

虽然赶不上目前其他主流生物识别技术的准确率,但这已经能证明心脏特征是一种有效实用的认证方式,未来将设备和算法调优后能达到更高的精准度。


Xu says the radiation from the device isn't harmful—just a fraction of that we get from our smartphones. But at a couple inches in size, the unit's not quite ready to fit inside a smartphone. "Uh, not yet. So we're waiting for the call from Apple." Because ya gotta have heart. But you also gotta have funding.


该设备释放出来的雷达信号,对人体的危害程度比WiFi、手机信号还小,并不用担心。

图片来源于@浅黑科技


LearnAndRecord

2015年2月8日

2017年10月11日

第977天

每天持续行动学外语

您可能也对以下帖子感兴趣

文章有问题?点此查看未经处理的缓存