Tuesday, July 5, 2022
SCIENMAG: Latest Science and Health News
No Result
View All Result
  • Login
  • HOME PAGE
  • BIOLOGY
  • CHEMISTRY AND PHYSICS
  • MEDICINE
    • Cancer
    • Infectious Emerging Diseases
  • SPACE
  • TECHNOLOGY
  • CONTACT US
  • HOME PAGE
  • BIOLOGY
  • CHEMISTRY AND PHYSICS
  • MEDICINE
    • Cancer
    • Infectious Emerging Diseases
  • SPACE
  • TECHNOLOGY
  • CONTACT US
No Result
View All Result
Scienmag - Latest science news from science magazine
No Result
View All Result
Home SCIENCE NEWS Technology and Engineering

Skoltech team shows how Turing-like patterns fool neural networks

March 11, 2021
in Technology and Engineering
0
Share on FacebookShare on Twitter

Skoltech researchers were able to show that patterns that can cause neural networks to make mistakes in recognizing images are, in effect, akin to Turing patterns found all over the natural world. In the future, this result can be used to design defenses for pattern recognition systems currently vulnerable to attacks. The paper, available as an arXiv preprint, was presented at the 35th AAAI Conference on Artificial Intelligence (AAAI-21).

Deep neural networks, smart and adept at image recognition and classification as they already are, can still be vulnerable to what’s called adversarial perturbations: small but peculiar details in an image that cause errors in neural network output. Some of them are universal: that is, they interfere with the neural network when placed on any input.

These perturbations can represent a significant security risk: for instance, in 2018, one team published a preprint describing a way to trick self-driving vehicles into “seeing” benign ads and logos on them as road signs. The fact that most known defenses a network can have against such an attack can be easily circumvented exacerbates this problem.

Professor Ivan Oseledets, who leads the Skoltech Computational Intelligence Lab at the Center for Computational and Data-Intensive Science and Engineering (CDISE), and his colleagues further explored a theory that connects these universal adversarial perturbations (UAPs) and classical Turing patterns, first described by the outstanding English mathematician Alan Turing as the driving mechanism behind a lot of patterns in nature, such as stripes and spots on animals.

The research started serendipitously when Oseledets and Valentin Khrulkov presented a paper on generating UAPs at the Conference on Computer Vision and Pattern Recognition in 2018. “A stranger came by and told us that this patterns look like Turing patterns. This similarity was a mystery for several years, until Skoltech master students Nurislam Tursynbek, Maria Sindeeva and PhD student Ilya Vilkoviskiy formed a team that was able to solve this puzzle. This is also a perfect example of internal collaboration at Skoltech, between the Center for Advanced Studies and Center for Data-Intensive Science and Engineering,” Oseledets says.

The nature and roots of adversarial perturbations are still mysterious for researchers. “This intriguing property has a long history of cat-and-mouse games between attacks and defenses. One of the reasons why adversarial attacks are hard to defend against is lack of theory. Our work makes a step towards explaining the fascinating properties of UAPs by Turing patterns, which have solid theory behind them. This will help construct a theory of adversarial examples in the future,” Oseledets notes.

There is prior research showing that natural Turing patterns – say, stripes on a fish – can fool a neural network, and the team was able to show this connection in a straightforward way and provide ways of generating new attacks. “The simplest setting to make models robust based on such patterns is to merely add them to images and train the network on perturbed images,” the researcher adds.

###

Media Contact
Ilyana Zolotareva
[email protected]
https://www.skoltech.ru/en/2021/03/skoltech-team-shows-how-turing-like-patterns-fool-neural-networks/

Tags: Computer ScienceResearch/DevelopmentRobotry/Artificial IntelligenceSystem Security/HackersTechnology/Engineering/Computer Science
Share25Tweet16Share4ShareSendShare
  • The COVID virus invading a cell

    COVID-19 fattens up our body’s cells to fuel its viral takeover

    97 shares
    Share 39 Tweet 24
  • Alcohol changes brain activity differently in male and female mice

    66 shares
    Share 26 Tweet 17
  • Scientists discover cancer trigger that could spur targeted drug therapies

    65 shares
    Share 26 Tweet 16
  • Scientists discover key to hepatitis A virus replication, show drug effectiveness

    65 shares
    Share 26 Tweet 16
  • Study shows convalescent plasma doesn’t benefit severely ill patients hospitalized with COVID-19

    65 shares
    Share 26 Tweet 16
  • New imaging technology less accurate than MRI at detecting prostate cancer, trial shows

    66 shares
    Share 26 Tweet 17
ADVERTISEMENT

About us

We bring you the latest science news from best research centers and universities around the world. Check our website.

Latest NEWS

nTIDE May 2022 COVID Update: Uncertainty about inflation tempers good news for people with disabilities

COVID-19 fattens up our body’s cells to fuel its viral takeover

The pair of Orcas deterring Great White Sharks – by ripping open their torsos for livers

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 190 other subscribers

© 2022 Scienmag- Science Magazine: Latest Science News.

No Result
View All Result
  • HOME PAGE
  • BIOLOGY
  • CHEMISTRY AND PHYSICS
  • MEDICINE
    • Cancer
    • Infectious Emerging Diseases
  • SPACE
  • TECHNOLOGY
  • CONTACT US

© 2022 Scienmag- Science Magazine: Latest Science News.

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
Posting....