Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Images altered to trick machine vision can influence humans too

deepmind.google

Post reply on HN
Print view View original

61–70 of 86 posts

Re: Images altered to trick machine vision can influence humans too

#61

Post by pixl97 »

sweezyjeezy wrote: post #10
Habgdnv wrote: post #7

I copied the image on the right into all possible AIs that i found on the net. They all told the that it is a vase with flowers. Even the most primitive. Something seems off here. Maybe they trained a model that is able to see "hidden" patterns and then they found that they can influence its mind with hidden patterns. For the rest of the general population (both humans and AIs) both images are the same.

I'm not sure that screen-shotting the image will work FWIW - any rescaling interpolation in rendering the image on the page or loading it for a model will likely reduce or nullify the effect. Also these perturbation based adversarial attacks are often model specific. You take the model's gradient at each pixel and iteratively perturbate the image to make it more and more confident that it's e.g. a cat.

So what you're saying run your images via a filter/resizer before feeding them into your AI.

Re: Images altered to trick machine vision can influence humans too

#62

Post by ndriscoll »

refulgentis wrote: post #51
Terr_ wrote: post #32

Earlier quoted context omitted.

To me that evokes a dimension of disbelief, or suspicion that the data is wrong, separate from what it does or doesn't resemble. Consider the difference between a human stating "that's an impossible nonsense picture, but if had to describe it then it's a half-Cat and half-Truck abomination" compared to a computer yielding " There is a 50% chance that is a Truck, and a 50% chance that is a Cat."

It might not be clear from the article, it took me a bit of scanning back and double checking myself: the impossible nonsense picture is _overlaid_ on the left hand picture to produce the right hand picture which is, at least to our conscious verbalizing mind, ~indistinguishable from the left hand picture. the interesting part is they _are_ distinguishable. as you point out, you'd expect a human to just be like "uhhh…

I don't think it is subconscious. In the bottom left of the vase picture, the "cat-like" version has "whiskers". The egg has a sort of caty blob on the right (though it's hard for me to decide whether the cat or truck version is more pronounced). The street picture has a white blob in the center that's "sheepier" than the chair version.

To me it seems like a conscious "if I had to pick the more cat-like one, this picture has a bit right here that sort of looks cat-like".

i.e. it's weird to say the images "influence" us. It seems more accurate to say that when prompted, we can perceive/understand what is confusing the model.

Re: Images altered to trick machine vision can influence humans too

#63

Post by Terr_ »

ndriscoll wrote: post #62
refulgentis wrote: post #51

Earlier quoted context omitted.

It might not be clear from the article, it took me a bit of scanning back and double checking myself: the impossible nonsense picture is _overlaid_ on the left hand picture to produce the right hand picture which is, at least to our conscious verbalizing mind, ~indistinguishable from the left hand picture. the interesting part is they _are_ distinguishable. as you point out, you'd expect a human to just be like "uhhh…

I don't think it is subconscious. In the bottom left of the vase picture, the "cat-like" version has "whiskers". The egg has a sort of caty blob on the right (though it's hard for me to decide whether the cat or truck version is more pronounced). The street picture has a white blob in the center that's "sheepier" than the chair version. To me it seems like a conscious "if I had to pick the more cat-like one, this pic…

In the case of giving people to look at subtly-distorted images and asking them "truck-vs-cat" and "chair-vs-sheep", I think it's likely people thought: "Huh, I guess one of those involves artificially constructed angles and the other tends to organic rounded bits, so I'll decide on which seems extra-present."

To me that seems a reasonable--and not very exciting--explanation for why humans may get similar this-or-that answers to a machine-vision model.

Re: Images altered to trick machine vision can influence humans too

#64

Post by trhway »

merelysounds wrote: post #25

All examples (sheep vs chair, dog vs bottle, cat vs truck, elephant vs clock) are organic vs inorganic. Perhaps participants are reusing bouba/kiki[1] skills, evaluating whether the image looks organic (rounded) or inorganic (spiky) - and making their choice accordingly. [1]: https://en.wikipedia.org/wiki/Bouba/kiki_effect

doesn't seem so. I think the key is probably here

>perturbed by a seemingly random pattern

ie. it doesn't really seem like a random pattern. It looks more like a match pattern taken from the mid-layers of a visual NN, the pattern of edges and vertexes which "lights up" more frequently when a cat image is presented to the NN. That would explain why an NN (even another one as they all converge on processing edges and vertexes) would mistake that image of vase for cat - in the altered image of vase those "cat" pattern edges and vertexes are present, just at the attenuated 2 levels amplitude, and the first layers of NN (usually converged to Gabors and Gaussian pulling like in the biological visuals cortex) would still detect those edges and will send it to the mid layers where it will "light up" similar "cat" patterns (it would be very illustrative if the authors provided the mixed-in image subjected to edge detection - that should clearly show a mix of original vase image edges and the added "cat" pattern ones). That also explains why people would point to such an image as more cat-like - the first layers and even the first-mid layers in our visual cortex are very similar to what the first and the mid layers of a well-trained NN converge to, and those our layers would similarly detect and "light up" to those mixed-in cat pattern edges and vertexes - we are just better at signal/noise, ie. clearly recognizing the vase and thus suppressing the "noise" of those cat patterns triggered in our visual cortex, so we wouldn't misclassify the image as a cat, yet we do feel something cat-like. Would be interesting to give that test to the people under hallucinogens, i.e. when the visual cortex mid-layers signals are subjected to much less frontal processing and the "noise" rises to the level of "signal".

Re: Images altered to trick machine vision can influence humans too

#65

Post by hytfyiv3j »

wizzwizz4 wrote: post #3

I don't see how this is surprising. The noise pattern in the first figure looks cat-like: I can see the ears, the head, the paws, the front half of the body… Having that “seemingly random pattern” to trace over would probably let me sketch a cat, something I can't normally do without a reference. (Though, the face is muddled and in the wrong place – almost like it's a cat collage – so I might only get the outline of…

If you actually went through with it, traced over that picture and posted it back to the thread that would be nice, as it seems not everyone can see it as clearly as you can.

Maybe even the original authors would appreciate it.

Re: Images altered to trick machine vision can influence humans too

#66

Post by glitchc »

riwsky wrote: post #44
glitchc wrote: post #11

This is a poor bit of research. The question "is it more cat-like?" Is leading as it specifically instructs the participant to look for cat-like features. The experimenters neglect to establish the null hypothesis.

The null hypothesis is that the participants are just as able to find those cat-like features in either perturbation of the image, and would pick the “right” one only 50% of the time.

No, the null hypothesis is that neither image elicits a cat-like response. It requires asking an open-ended question such as "what does this image look like to you?" Once you prime the subject, you have artificially restricted the responses to "Not really, sure, kinda?"

Remember, the ML model is objectively selecting cat with (very) high probability out of the entire corpus of possible responses. The human should be given the same range of possible responses to objectively establish bias. Since no human would say 'cat-like' for any of those images, it suggests a fairly large gap between human and machine perception. We've got a long road ahead of us.

Re: Images altered to trick machine vision can influence humans too

#67

Post by EnigmaFlare »

It would be amazing if NNs can eventually create art that inspires emotions in the audience without them realizing how. That seems to be a feature of normal art but maybe machines can one day do better than the best human artists.

Unfortunately, the researchers will probably never follow up on their work and improve things. That's the sad story of most fun research, they just give up and forget about it once it's published and nobody else seems to want to do the work themselves. Though perhaps that's because most research doesn't actually have any potential for improvement or is false to begin with and the authors know it.

Re: Images altered to trick machine vision can influence humans too

#69

Post by sweezyjeezy »

pixl97 wrote: post #61
sweezyjeezy wrote: post #10

Earlier quoted context omitted.

I'm not sure that screen-shotting the image will work FWIW - any rescaling interpolation in rendering the image on the page or loading it for a model will likely reduce or nullify the effect. Also these perturbation based adversarial attacks are often model specific. You take the model's gradient at each pixel and iteratively perturbate the image to make it more and more confident that it's e.g. a cat.

So what you're saying run your images via a filter/resizer before feeding them into your AI.

haha that's an interesting point! Recompress as a jpeg maybe.

Re: Images altered to trick machine vision can influence humans too

#70

Post by krisoft »

EnigmaFlare wrote: post #67

It would be amazing if NNs can eventually create art that inspires emotions in the audience without them realizing how. That seems to be a feature of normal art but maybe machines can one day do better than the best human artists. Unfortunately, the researchers will probably never follow up on their work and improve things. That's the sad story of most fun research, they just give up and forget about it once it's pub…

> Unfortunately, the researchers will probably never follow up on their work

That is a very weird assumption.

> nobody else seems to want to do the work themselves

If you are so interested why don’t you try your hand at it? This particular research is super easy and accessible.

Post reply on HN