Images altered to trick machine vision can influence humans too

Re:Images altered to trick machine vision can influence huma... by pbhjpbhj on Tuesday January 02, 2024 @07:35PM (#38849175)

Caption on the first image :

>when perturbed by a seemingly random pattern across the entire picture (middle), with the intensity magnified for illustrative purposes

I don't understand? The pattern is not "seemingly random", it is "seemingly chosen to have subtle cat-features". One sees the ears at the top of them image and face-like features below.

So, is it "we perturbed images to overlay cat-like features on a visual level that humans don't generally perceive but ML models were able to perceive; and then ML models perceived them"?

Can someone précis the results and why they're interesting because on the face of it this seems like a very obvious outcome?

Do I need to make a new year resolution to actually read the articles?

[ Parent | on HN ]

Re:Images altered to trick machine vision can influence huma... by refulgentis on Tuesday January 02, 2024 @07:39PM (#38849211)

You're not wrong, but this is interesting insomuch as it's it's more from 100% bug on the bug-feature axis to...well, wetware has this bug too it's just subconcious.

[ Parent | on HN ]

Re:Images altered to trick machine vision can influence huma... by wizzwizz4 on Tuesday January 02, 2024 @07:46PM (#38849278)

But it's not a bug. You look at it and go “hey, those flowers look like a cat”, in the same way you go “that cloud looks like a horse” or “that tree looks like a face” (though, many people have specialised human face processing machinery in their visual systems, so this last example is potentially a little different). It's not a misclassification, just an awareness.

[ Parent | on HN ]