New attack on neural networks can alter the purpose of the neural network.

A surprising adversarial attack, whereby a perturbation to all input images can "reprogram" a poorly-defended neural network to change its task entirely. e.g. turn an ImageNet classifier into a network that counts squares.

Arxiv: https://arxiv.org/pdf/1806.11146.pdf

#Goodfellow #gbrain #adversarialattack

Ian Goodfellow (Google Brain) will have a free webinar on Adversarial Machine Learning

https://event.on24.com/eventRegistration/EventLobbyServlet?target=reg20.jsp&partnerref=twitterShareFromReg&ms=1531251848082&eventid=1633807&sessionid=1&key=8B7A8F4B65B54C35752F8A6FE23F641A&regTag=&sourcepage=register

#webinar #google #gbrain