Live data from Hacker News

The Problem with Perceptual Hashes

rentafounder.com

91–100 of 440 posts

Re: The Problem with Perceptual Hashes

#91

Earlier quoted context omitted.

I don't even have nude photos of my wife. The only person who might would be the NSA contractor assigned to watch her.

what function does the word "even" perform in this sentence?

I believe it's meant to be "Even _I_ don't have..."

Re: The Problem with Perceptual Hashes

#92

Earlier quoted context omitted.

> Do you really want a false positive to result in your intimate moments getting shared around some outsourced boiler room for laughs? these people also have no incentive to find you innocent for innocent photos. If they err on the side of false-negative, they might find themselves at the wrong end of a criminal search ("why didn't you catch this"), but if they false-positive they at worse ruin a random person's life…

Even still this has to go to the FBI or other law enforcement agency, then it’s passed on to a prosecutor and finally a jury will evaluate. I have a tough time believing that false positives would slip through that many layers. That isn’t to say CASM scanning or any other type of drag net is OK. But I’m not concerned about a perceptual hash ruining someone’s life, just like I’m not concerned about a botched millimete…

> But I’m not concerned about a perceptual hash ruining someone’s life

I want ZERO computerized algorithms involved in any law enforcement process - especially the "criminal hunting" steps.

Re: The Problem with Perceptual Hashes

#93

Earlier quoted context omitted.

Rookie mistake. Three rules to live by: 1) Always pay your taxes 2) Don’t talk to the police 3) Don’t take photographs with your clothes off

I might amend #2 a bit to read "Be friends with the police" as that has historically been more beneficial to those who are.

Lots of people have believed that they were friends with the police and were actually being manipulated into metaphorically hanging themselves- some of them innocent.

Counterargument, why you should not talk to the police (In the US): https://youtu.be/d-7o9xYp7eE

Re: The Problem with Perceptual Hashes

#94

> an Apple employee will then look at your (flagged) pictures. This means that there will be people paid to look at child pornography and probably a lot of private nude pictures as well.

And what do you think the content moderation teams employed by Facebook, YouTube, et al. do all day?

There's a big difference in the expectation of privacy between what someone posts on "Facebook, Youtube, et al" and what someone takes a picture of but doesn't share.

Re: The Problem with Perceptual Hashes

#95

Earlier quoted context omitted.

And what do you think the content moderation teams employed by Facebook, YouTube, et al. do all day?

They look at content that people actively and explicitly chose to share with wider audiences.

While that's a snappy response, it doesn't seem to have much to do with the concern about perverts getting jobs specifically to view child abuse footage, which is what I thought this thread was about.

Re: The Problem with Perceptual Hashes

#96

I've also implemented perceptual hashing algorithms for use in the real world. Article is correct, there really is no way to eliminate false positives while still catching minor changes (say, resizing, cropping, or watermarking). I'm sure I'm not the only person with naked pictures of my wife. Do you really want a false positive to result in your intimate moments getting shared around some outsourced boiler room for…

> Do you really want a false positive to result in your intimate moments getting shared around some outsourced boiler room for laughs? these people also have no incentive to find you innocent for innocent photos. If they err on the side of false-negative, they might find themselves at the wrong end of a criminal search ("why didn't you catch this"), but if they false-positive they at worse ruin a random person's life…

This is mostly orthogonal to the author's original point (with which I concur, having also implemented image similarity via hashing and hamming distance). There just aren't a lot of knobs to tune using these algorithms so it's difficult if not impossible to make small changes to err on the side of reducing false positives.

Re: The Problem with Perceptual Hashes

#97

Given all the zero day exploits on iOS I wonder if it's now going to be viable to hack someone's phone and upload child porn to their account. Apple with happily flag the photos and then, likely, get those people arrested. Now they have to, in practice, prove they were hacked which might be impossible. Will either ruin their reputation or put them in jail for a long time. Given past witch hunts it could be decades be…

Wouldn't this risk exist already, as long as it is uploaded to icloud?

Re: The Problem with Perceptual Hashes

#98

Earlier quoted context omitted.

> Do you really want a false positive to result in your intimate moments getting shared around some outsourced boiler room for laughs? these people also have no incentive to find you innocent for innocent photos. If they err on the side of false-negative, they might find themselves at the wrong end of a criminal search ("why didn't you catch this"), but if they false-positive they at worse ruin a random person's life…

Does claiming a false positive not run the risk of libel?

I doubt it. The claim isn't being published and you'd have a hard time proving damages.

Re: The Problem with Perceptual Hashes

#99
post #80
post #78

Earlier quoted context omitted.

So what are we going to do about it? I have a large user base on iOS. Considering a blackout protest.

IMHO, Unless everything being E2E encrypted becomes the law we can’t do anything about it because that’s not Apple’s initiative but comes from people whose job is to know things and they cannot resist keeping their hands out of these data collecting devices. They promise politicians that all the troubles will go away if we do that. Child pornography, Terrorism? Solve it the old way. I don’t know why citizens are obli…

Unless I missed a big part of this story Apple isn't being compelled by a court to comply - so if the presence of this tech causes a larger PR stink than publicly backing out of the rollout then Apple will read the writing on the wall.

Public and political pressure is definitely an issue - but it's still soft-pressure so applying more pressure in the other direction will be compelling to Apple.

Re: The Problem with Perceptual Hashes

#100
post #75

I agree with the article in general except part of the final conclusion > The simple fact that image data is reduced to a small number of bits leads to collisions and therefore false positives Our experience with regular hashes suggests this is not the underlying problem. SHA256 hashes have 256 bits and still there are no known collisions, even with people deliberately trying to find them. SHA-1 only has only 160 bit…

This isn’t due to the entropy of the hash but due to the entropy of the source data.

These algos work by limiting the color space of the photo, usually to only black and white (not even grey scale) resizing it to a fraction of its original size and then chopping it into tiles using a fixed size grid.

This increases the chances of collisions greatly because photos with a similar composition are likely to match on a sufficient number of tiles to flag the photo as a match.

This is why the women image was matched to the butterfly image, if you turn the image to B&W resize it to something like 256x256 pixels and divide it into a grid of say 16 tiles all of a sudden a lot of these tiles can match.

Post reply on HN