[HN Gopher] Images altered to trick machine vision can influence...
       ___________________________________________________________________
        
       Images altered to trick machine vision can influence humans too
        
       Author : xnx
       Score  : 31 points
       Date   : 2024-01-02 20:30 UTC (2 hours ago)
        
 (HTM) web link (deepmind.google)
 (TXT) w3m dump (deepmind.google)
        
       | xnx wrote:
       | Though not mentioned in the blog post, this seems like it would
       | have some applications for true "subliminal" advertising.
        
         | tudorw wrote:
         | 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 want.
        
           | stefs wrote:
           | If you can't afford it there's SafeSight Free*, which only
           | allows a selection of sponsors deemed acceptable by our
           | ethics/advertising committee.
        
       | wizzwizz4 wrote:
       | 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 a cat.)
       | 
       | This study's result is implied by the stronger result: "a neural
       | network's notion of a category sometimes resembles members of the
       | category". I'm sure a competent sketch artist could yield similar
       | or better results, being able to take advantage of peculiarities
       | of the human visual system. (In fact, that might be a good
       | follow-up study: I might claim it if nobody else does.)
        
         | falcolas wrote:
         | Maybe my brain's just odd... but I don't see the cat-like
         | figure in the noise. I just see noise. There are a few edges,
         | but nothing really cat-like to my mind.
         | 
         | My subconscious pattern recognition for faces and such has
         | always been weak, fwiw.
        
       | graypegg wrote:
       | One thing I'd be curious about is how many of these perturbations
       | can you stack up?
       | 
       | A picture of a cat still retains an obvious cat-ness to humans,
       | even when it's been tainted by "truck-ness", but they're slightly
       | more likely to agree it's more truck-y. That response scales with
       | how perturbed it is. (Fig 3 in paper.) If you added the
       | perturbations for "vehicle-ness", would that response be stronger
       | while affecting the image less than cracking up the intensity on
       | the effect? Could you start combining separate concepts, and pick
       | them out individually as like... "activation scores" or
       | something?
       | 
       | If so, feels like compression all of a sudden. I know there's a
       | ton of other ML compression things out there, but that just feels
       | like it could be really information dense.
        
       | Habgdnv wrote:
       | 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.
        
         | graypegg wrote:
         | Compression possibly? I'm seeing .webps on the paper. Putting
         | them into webpinfo gives me:                   File:
         | /.../41467_2023_40499_Fig3_HTML.png.webp         RIFF HEADER:
         | File size: 671810         Chunk VP8  at offset     12, length
         | 671798           Width: 2000           Height: 2255
         | Alpha: 0           Animation: 0           Format: Lossy (1)
         | No error detected.
         | 
         | So since they're lossy, maybe the subtly is lost?
         | 
         | Edit: The image on the article itself is an SVG, containing 3
         | jpegs. So that's absolutely mangled in comparison to the
         | paper's lossy images.
         | 
         | https://deepmind.google/api/blob/website/images/Figure0_svg....
        
         | sweezyjeezy wrote:
         | 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.
        
       | glitchc wrote:
       | 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.
        
         | dr_dshiv wrote:
         | 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.
        
       | paxys wrote:
       | Very interesting results. There's a massive overlap between the
       | current generation of AI research/development and neuroscience,
       | and it's fitting that by so desperately trying to make a computer
       | intelligent we are unexpectedly learning more about how our own
       | brains work.
        
         | nomel wrote:
         | > we are unexpectedly learning
         | 
         | Many expected this, decades ago. But, also, many claimed that
         | neural networks have nothing to do with the brain. I think
         | we're slowly inching towards and understanding that we're the
         | result of some fundamentals of information organization, and
         | those fundamentals are realized in biology, rather than come
         | from it. Those fundamentals are now showing themselves in
         | silicon.
        
       | TehShrike wrote:
       | In 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
        
       ___________________________________________________________________
       (page generated 2024-01-02 23:00 UTC)