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

Making Visible Watermarks More Effective

research.googleblog.com

Post reply on HN
Print view View original

41–50 of 98 posts

Re: Making Visible Watermarks More Effective

#41

Post by UnoriginalGuy »

FTA wrote: post #19

This reminds me a bit of the common argument for locks: it keeps regular people honest. Watermarks are designed really to deter you from casually copying an image and pasting it on your site, unless you don't care at all about a watermark showing (e.g. a blog site with a few followers). I'm sure a mixture of the original computer vision technique plus some smearing on the original image could wash away even these ran…

Indeed.

But the practice has somewhat backfired. Some members of the general public now erroneously believe that any image lacking a watermark is a public works.

For one specific anecdote, a local popular bar lifted an image of their establishment found on Flickr and used it in their marketing materials (posters mostly). When contacted about this by the photographer, the establishment claimed due to the lack of watermark and being their establishment they were well within their rights to use the photograph. They eventually agreed to stop using it as a "gesture of goodwill" (but never paid for it, and weren't sued as the cost of even small claims was likely higher than the damages).

You read stories like this often if you hang out on photography forums. The general public seem to conflate a watermark with copyright and a lack of watermark with a public works. Basic copyright needs to be taught in schools.

Re: Making Visible Watermarks More Effective

#42

Post by Fiahil »

0942v8653 wrote: post #16
bwang29 wrote: post #11

I think the question is why would we need protective watermark anyway if stock photo companies are already crawling and sometimes phishing for use of licensed stock photography on the web and then directly send out an DCMA or a charge? I've been in many situations where the copyright owners reached out for damage fees after downloading a full-res, un-watermarked photo from free stock photos sites in blog posts, so I'…

What if you put it into a video? Photoshop it into a meme? Drop it in to an internal presentation? These are all situations where the copyright of a stock image owner is infringed and yet there is very little that automated processes can do to detect them.

What is the issue with dropping copyrighted images into internal presentations ?

Re: Making Visible Watermarks More Effective

#43

Post by pps43 »

Retric wrote: post #28
pps43 wrote: post #9

Earlier quoted context omitted.

If the same watermark is added to each image, then you can estimate the watermark by averaging many images and subtract it to get original image back. If each image has a slightly different watermark, then simple averaging won't work. Instead you need to come up with a model that describes how the watermark is changing, then estimate parameters of that model. The more complex the change, the more images you need for…

Averaging will still provide the average watermark given a large enough sample set. Which is then a single geometric transformation from the used watermark. Estimating that transformation for a specific image should be fairly straightforward as they could easily reverse a watermark that was offset. At which point it's reversible.

Not really, because in each image parts of the watermark will be in different places. You don't even need to warp the watermark to achieve that, it's enough to randomly reposition it.

Re: Making Visible Watermarks More Effective

#44

Post by euyyn »

UnoriginalGuy wrote: post #41
FTA wrote: post #19

This reminds me a bit of the common argument for locks: it keeps regular people honest. Watermarks are designed really to deter you from casually copying an image and pasting it on your site, unless you don't care at all about a watermark showing (e.g. a blog site with a few followers). I'm sure a mixture of the original computer vision technique plus some smearing on the original image could wash away even these ran…

Indeed. But the practice has somewhat backfired. Some members of the general public now erroneously believe that any image lacking a watermark is a public works. For one specific anecdote, a local popular bar lifted an image of their establishment found on Flickr and used it in their marketing materials (posters mostly). When contacted about this by the photographer, the establishment claimed due to the lack of water…

Probably just started using a different picture without permission either :/

Re: Making Visible Watermarks More Effective

#45

Post by cs702 »

The researchers identify the watermarks to be removed by finding pixel patterns that persist across a large number of images, as shown on this animation: https://1.bp.blogspot.com/-cJwNoUxIBzM/WZTDpw3ru6I/AAAAAAAAB... . Their proposed solution is to randomly warp the watermarks.

Unfortunately, their solution could be quickly defeated with image-to-image generative adversarial convnets trained to... remove watermarks from image pairs. (That is, instead of training a model to change, say, image style or resolution, train it to remove artificially added watermarks.)

Re: Making Visible Watermarks More Effective

#46

Post by Kdparker »

azdle wrote: post #34
jotato wrote: post #27

Earlier quoted context omitted.

i mean to remove the watermark on that specific image

The idea is to use a randomly-warped per-image watermark. So in theory, there's exactly one distinct watermark per image, so 1000 of the same image would have the same watermark.

It would be easy to change that hash by modifying one pixel per image (pick one that is unaffected by the watermark).

Re: Making Visible Watermarks More Effective

#47

Post by gknoy »

Fiahil wrote: post #42
0942v8653 wrote: post #16

Earlier quoted context omitted.

What if you put it into a video? Photoshop it into a meme? Drop it in to an internal presentation? These are all situations where the copyright of a stock image owner is infringed and yet there is very little that automated processes can do to detect them.

What is the issue with dropping copyrighted images into internal presentations ?

If I put a copyrighted image into an internal presentation, I am still copying and redistributing this image. Without permission to do so, that seems like it's straight-up infringement. I don't believe it usually meets the criteria for fair use, either.

Re: Making Visible Watermarks More Effective

#48

Post by azdle »

Kdparker wrote: post #46
azdle wrote: post #34

Earlier quoted context omitted.

The idea is to use a randomly-warped per-image watermark. So in theory, there's exactly one distinct watermark per image, so 1000 of the same image would have the same watermark.

It would be easy to change that hash by modifying one pixel per image (pick one that is unaffected by the watermark).

I'm not sure what you're suggesting. Change what hash? Of the image? I can't think of how that could help with removing a watermark.

Re: Making Visible Watermarks More Effective

#50

Post by throwaway2016a »

FTA wrote: post #19

This reminds me a bit of the common argument for locks: it keeps regular people honest. Watermarks are designed really to deter you from casually copying an image and pasting it on your site, unless you don't care at all about a watermark showing (e.g. a blog site with a few followers). I'm sure a mixture of the original computer vision technique plus some smearing on the original image could wash away even these ran…

Whenever they tell me my smart lock can be hacked I respond... "yeah, and someone can also just take a brick, smash my window, and go through that way."
Post reply on HN