Easily Fooling Deep Neural Networks
Data Skeptic16 Jan 2015

Easily Fooling Deep Neural Networks

My guest this week is Anh Nguyen, a PhD student at the University of Wyoming working in the Evolving AI lab. The episode discusses the paper Deep Neural Networks are Easily Fooled [pdf] by Anh Nguyen, Jason Yosinski, and Jeff Clune. It describes a process for creating images that a trained deep neural network will mis-classify. If you have a deep neural network that has been trained to recognize certain types of objects in images, these "fooling" images can be constructed in a way which the network will mis-classify them. To a human observer, these fooling images often have no resemblance whatsoever to the assigned label. Previous work had shown that some images which appear to be unrecognizable white noise images to us can fool a deep neural network. This paper extends the result showing abstract images of shapes and colors, many of which have form (just not the one the network thinks) can also trick the network.

Det här avsnittet är hämtat från ett öppet RSS-flöde och publiceras inte av Podme. Det kan innehålla reklam.

Avsnitt(601)

Reducing the Impact of Ship Noise on Marine Mammals

Reducing the Impact of Ship Noise on Marine Mammals

Human shipping operations have increased significantly in the past few decades.  While that means international trade and cheap goods for humans, it also means the ocean has experienced an increase in...

1 Juli 202436min

Analysis of Unstructured Data

Analysis of Unstructured Data

Robbie Moon from the Georgia Tech Scheller College of Business joins us to discuss the analysis of unstructured data and the application of NLP methodologies towards financial data.

28 Juni 202427min

iNaturalist

iNaturalist

Have you ever participated in citizen science? Do you want to? One of the most popular platforms for crowdsourcing biodiversity data is iNaturalist. In addition to being a great science tool, the iNat...

24 Juni 202437min

Learn to Code

Learn to Code

Do you code or are you interested in learning to code? Join us today and hear from three individuals that are at very different stages of their coding journeys. Becky Hansis-O'Neill (also our co-host ...

18 Juni 202449min

Animal Computer Interaction

Animal Computer Interaction

You've heard of Human Computer Interaction (HCI), now get ready for Animal Computer Interaction (ACI). Ilyena has made a career developing computer interfaces for non-human animals. She has worked wit...

10 Juni 202442min

Ape Gestures

Ape Gestures

Cat observes great apes in the wild and in the lab to crack the code of their gestural communication. We discussed the challenges and benefits of studying apes in the wild vs in the lab. Cat also shar...

3 Juni 202449min

Evaluating AI Abilities

Evaluating AI Abilities

In this episode, Kozzy discusses his endeavors to compare the cognitive abilities of humans, animals, and AI programs. Specifically, we discussed object permanence, the ability to understand an object...

27 Maj 202449min

HMMs for Behavior

HMMs for Behavior

Théo Michelot has made a career out of tackling tough ecological questions using time-series data. How do scientists turn a series of GPS location observations over time into useful behavioral data? G...

20 Maj 202445min

Populärt inom Vetenskap

p3-dystopia
dumma-manniskor
allt-du-velat-veta
kapitalet-en-podd-om-ekonomi
sexet
rss-ufobortom-rimligt-tvivel
rss-vetenskapsradion
svd-nyhetsartiklar
medicinvetarna
rss-vetenskapsradion-2
det-morka-psyket
ufo-sverige
rss-spraket
hacka-livet
paranormalt-med-caroline-giertz
halsorevolutionen
barnpsykologerna
ideer-som-forandrar-varlden
rss-klotet
pojkmottagningen