This is one of the most nonsensical things I've read for a long time.
Images altered to trick machine vision can influence humans too
21–30 of 86 posts
Re: Images altered to trick machine vision can influence humans too
#22I 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.
Adversarial examples depend strongly on the classifier used because the researchers are directly analyzing the responses of the network. You would need to use the same model.
This is also why the “anti-facial recognition” shirts are silly. If they had adversarial noise at all, you likely won’t know about or have access to the model you’re trying to fool.
Re: Images altered to trick machine vision can influence humans too
#23"In our example, we may see a vase of flowers, but some activity in the brain informs us there’s a hint of cat about it."
IMHO this is not the same as computer vision thinking a rolled over school bus is a snow plow.
This is asking if someone sees an elephant or a unicorn in a cloud.
Asking if a picture of a stop light at an intersection is "cat like" seems to be pretty suspectable to over fitting.
Rorschach inkblot test is pretty much pseudoscience, can someone please explain how this is not similar?
Re: Images altered to trick machine vision can influence humans too
#24This 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.
They made stimuli to be either cat-like or sheep-like, for instance, and asked them to pick the more cat-like. It wasn’t between cat and nothing.
Cat-like for machines, not humans. What if they had asked which one is more butterfly like and then humans would pick the same one? Prompting humans for a cat is not without consequences - the machines were not prompted to find a cat.
Re: Images altered to trick machine vision can influence humans too
#25All 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.
Re: Images altered to trick machine vision can influence humans too
#26In case you were wondering what N was, their first experiment involved 16 undergrads psych students and the second experiment involved 12. https://link.springer.com/article/10.3758/BF03206939 Edit: I believe this linked survey is not the subject of the OP.
What do your power calculations for the effect size say a minimal sample should be?
My stats professor worked in cardiology and he shared a paper about n=1 or 2 studies: "If nobody dies, is everything OK?"
Re: Images altered to trick machine vision can influence humans too
#27we should code up a real online game to gather the stats here. maybe we'll find a real subliminal messaging model
Re: Images altered to trick machine vision can influence humans too
#28Why do they assume that 50% choose cat and 50% choose truck? Or did I miss the part where they show the participants an untampered version of the image to create a baseline. I mean, cats are small and cute, trucks are big and stinky. And wouldn't the subject of the image also make the participants possibly lean to one of the answers? Flowers in a vase go well with a cat (both can be found in an apartment, for example…
I think you misunderstand the experiment. They take an image of the flower, they perturb it so that the neural network classifies it as a "cat". They take another copy, perturb it so that the neural network classifies it as a "truck". They ask the subjects which one is more cat-like. A coin will choose the correct image 50% of the time. Likewise, a human that is not influenced by the pertubations will also pick correctly 50% of the time (as long as there is no systematic influence like the correct one is always on the right, the interviewer is unaware, etc.). What they found is that humans do pick the correct one more frequently than random.
Re: Images altered to trick machine vision can influence humans too
#29Though not mentioned in the blog post, this seems like it would have some applications for true "subliminal" advertising.
No problem, simply subscribe to our service SafeSight, using the latest imaging technology everything you see is processed by our AI powered manipulation sensing algorithms before your eyes see it, real-time screening and flagging of events together with defensive response suggestions let you take charge. Control your own mind. Feel safe looking out on a world full of messaging and do what you want, not what they wan…
Thanks, but I've already got a pair of Joo Janta 200 Super-Chromatics.
https://odysee.com/@NoVax:c/Hitchhikers-Guide-to-the-Galaxy-...
Re: Images altered to trick machine vision can influence humans too
#30Why do they assume that 50% choose cat and 50% choose truck? Or did I miss the part where they show the participants an untampered version of the image to create a baseline. I mean, cats are small and cute, trucks are big and stinky. And wouldn't the subject of the image also make the participants possibly lean to one of the answers? Flowers in a vase go well with a cat (both can be found in an apartment, for example…
I think you misunderstand the experiment. They take an image of the flower, they perturb it so that the neural network classifies it as a "cat". They take another copy, perturb it so that the neural network classifies it as a "truck". They ask the subjects which one is more cat-like. A coin will choose the correct image 50% of the time. Likewise, a human that is not influenced by the pertubations will also pick corre…
Not clear if they had asked for "butterfly" instead of "cat" if 50/50 would have been the result. Similarly, if random perturbations influence choice, the baseline should include the noise from that.