Categories: Video news

Programmers came up with an unusual method of hacking the phone. Video

Progressives simply create any sound.

Engineers from Zhejiang University have learned to create inaudible ultrasonic for people teams which voice assistants on smartphones and other devices perceive the same as human speech. To smartphone heard of the team, the ultrasonic transmitter must be close to it and not in a very noisy environment. The researchers also suggested a way to protect from this attack.

Voice assistants already have almost all the smartphones, the same technology is used in various centers, smart home control, such as Google Home or the Amazon Echo. They can perform a variety of tasks, including making phone calls, send messages or read them, take photos and video, and also have access to many other sensitive data. Because of this, in recent years, often appear messages about new methods of breaking such programs.

Chinese researchers decided to create a method of hacking that will be universal for all platforms, and thus invisible to the victim. So the attack was invisible, the engineers used the ultrasound. They modulated write commands so that the output signal has a frequency that cannot be perceived by the human ear. However, since almost all of the microphones in the smartphones contain low-pass filter, the researchers had to find a way to turn back the ultrasonic signal perceived by the microphone. For this, they used the inevitable nonlinear distortion in microphones and amplifiers, which created new frequencies already in the microphone and thus demodulator signal.

The researchers tested the method with several voice assistants. One of them is the Apple Siri was a bit more resistant to this attack due to the fact that he remembers his master’s voice and responds only to it.

Engineers offered to record a small snippet of the speech owner, to isolate desired sounds, and then form one team that a smartphone will take over the team’s owner. By the way, last year, Google created a neural network that can realistically simulate the speech of man that can also be used for such attacks.

See also: Scientists have invented an unusual method of communication with aliens

This method of hacking, there are many restrictions. For example, the attacker or his equipment with ultrasonic emitter should not be more than 1.75 meters from the goal. The engineers also tested the break-in at different levels of ambient noise, and unless the terms of office recognition was successful in the vast majority of cases, for a success rate of teams was reduced to 30 percent. As protection from attack by using ultrasound the researchers propose to use support vector machines, which will distinguish these commands from the commands spoken by the person.

© 2017 – 2019, paradox. All rights reserved.

paradox

Share
Published by
paradox

Recent Posts

Ukrainian government warns citizens of possible full mobilization

The entire Ukrainian society would need to make sacrifices and forget about their peaceful lives…

6 hours ago

PM Mishustin announces shake-up of Russian government

Russian Prime Minister Mikhail Mishustin on Saturday unveiled his first picks for the new cabinet,…

15 hours ago

Ukraine shells cafe in Russia’s Donetsk (VIDEO)

One person was killed and at least four others wounded in a Ukrainian rocket strike…

24 hours ago

Poke the bear and find out: Here’s why the West should finally listen to Russia’s warnings

We have been through an intense, if muffled crisis in the ongoing political-military confrontation between…

1 day ago

Russia launches new offensive – Kiev

Ukrainian President Vladimir Zelensky claimed on Friday that Russian forces have launched a new large-scale…

2 days ago

NATO soldiers operating in Ukraine – Polish PM

NATO soldiers are already in Ukraine helping Kiev but the US-led bloc does not want…

2 days ago