Another AI Fubar

itman · Apr 19, 2017

Adding Noise to Images Is Enough to Fool Google's Top-Notch Image Recognition AI

A team of computer experts from the Network Security Lab (NSL) at the University of Washington proved this week that a determined adversary could trick Google's Cloud Vision API into misclassifying submitted images.

This research comes as AI-based image classification systems are becoming more popular and are increasingly adopted by more online services, where they are used to catch and block images of abuse, adult, violent, or other inappropriate content before they are published on their live sites.

Despite working on top of a complex machine-learning-powered algorithm, the research team says it found a simple method of deceiving Google's Cloud Vision service.

Google Cloud Vision API vulnerable to trivial attack

Their attack relies on adding low quantities of "noise" to an image. The noise level varied from 10% to 30% but was enough to deceive Google's AI, while also keeping the images clear and visible for human observers.
Click to expand...

https://www.bleepingcomputer.com/ne...-fool-googles-top-notch-image-recognition-ai/

guest · Dec 1, 2018

Google's image recognition AI fooled by new tricks
November 30, 2018
https://www.zdnet.com/article/googles-best-image-recognition-system-flummoxed-by-fakes/

[...] researchers at Auburn University in Auburn, Ala., and media titan Adobe Systems. In a paper released this week, they showed that top image-recognition neural networks easily fail if objects are moved or rotated even by slight amounts.

A fire truck, for example, seen from head on, could be correctly recognized. But once pointed up in the air, and turned around several times, the same fire truck is mis-classified by the neural net as a school bus or a fireboat or a bobsled.
The upshot is that the state of the art in image recognition is "naive," and some greater understanding of three-dimensional structures seems needed to help them get better.

The paper, Strike (with) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects, is posted on the arXiv pre-print server; The authors have also posted the code they used to do it on GitHub.
It's clear neural networks might need some substantial help to move forward.
Click to expand...

Paper (PDF): https://arxiv.org/pdf/1811.11553.pdf

Log in or Sign up

Another AI Fubar

itman Registered Member

guest Guest

Log in or Sign up

Another AI Fubar

itman Registered Member

guest Guest

Useful Searches