Author Topic: proving video plagiarism with compression artifacts  (Read 2627 times)

0 Members and 1 Guest are viewing this topic.

Offline NiHaoMikeTopic starter

  • Super Contributor
  • ***
  • Posts: 9154
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
proving video plagiarism with compression artifacts
« on: April 16, 2024, 05:03:44 am »
How difficult (or even possible) would it be to prove that video B that looks identical to video A is in fact a plagiarized copy of A that has since been reencoded? (And not vice versa.) My understanding is that A would have the compression artifacts from its encoding process, while B would have compression artifacts from its own encoding as well as compression artifacts of A that didn't get masked.
I'm aware that there are "watermarking" algorithms specifically designed for that, my question is if the original compression artifacts can act as that watermark.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.

Offline Haenk

  • Super Contributor
  • ***
  • Posts: 1162
  • Country: de
Re: proving video plagiarism with compression artifacts
« Reply #1 on: April 16, 2024, 07:08:01 am »
I'd say that will work, in general.
However consider "AI enhancing" (or other enhancing), which might unblur, sharpen, add details (that might or might not be there in the first place), but I'm sure, with a bit of experience, one could identify this as well.
I remember my first DVD-player, using an upscaler ASIC from Faroudja, which essentially does the mentioned picture enhancing and re-encoding. Amazing, but certainly detectable, if you know, what to look for...

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo