Music

So, welcome to the multimedia security lecture.

We started to talk about watermarking last week, and this week I would like to continue

and put a little bit more emphasis on these criteria for good watermarks that we mentioned

last time.

So particularly robustness of watermarks, watermark security, and fidelity.

So the thing that a watermark is ideally imperceptible.

Okay.

But first of all, some details with respect to the organization.

So I think like after today, you essentially know enough passwords or something to just

make an attempt to screen the literature a little bit.

And I'll put, I will put, I think I will put three papers, two or three papers on Stutt

Pfeiffer, and I would like you to have a look at them.

You don't need to go through them line by line.

But it would be great if you could see, okay, can I grasp the core idea?

How can I sort these papers into what we've discussed so far?

Yeah, so how does it fit into that framework?

For example, what do they do with respect to robustness?

What do they do with respect to perceptual fidelity?

Things like that.

That would be questions that would be relevant in order to get an idea.

Okay, like how advanced is this work?

How solid is this work?

I mean, okay, hopefully they're all solid.

But okay, not to be able to argue about the content of the methods in actual publications.

So this is not staged or something.

Yeah, okay.

But on the other hand, if you, I don't know, if you look at an equation and it all doesn't

make sense or so, then don't worry.

Yeah, we can, I will come back to this next week then.

So I upload these papers tonight to Stuttgart and then you have a week to look at them.

And at the beginning of the lecture next week, I think we could just talk about them like

for half an hour or so and then move on.

Okay, so that was one thing.

The other thing is if you have the schedule in mind that I originally proposed for the

lecture, then you'll notice that I'm deviating a little bit.

Okay, so that was the draft for the schedule.

And deviating a little bit.

Oops, is that large enough, big enough?

So then we have, like if we are, next week, if we are also talking a little bit at least

about watermarking, you see that watermarking takes, I mean, fingerprinting is sort of the

watermarking thing, yeah, but that watermarking will take more space, a little bit more space

than anticipated.

I think that's fine.

It's an important topic.

And signal processing in the encrypted domain is maybe more like a fun topic, but it's not

super central to multimedia security, I'd say.

So we will have a little bit wait here and I'll just squeeze that in some corner or so.

And then we also had a request to make, do an exercise on steganography or steganalysis,

which is, I think it's a great idea, but it's something that I cannot accommodate by just