Welcome everybody. It's a great pleasure to have you all here for our pattern recognition symposium.

And I have a couple of announcements to start with. So we have in total 39 public talks.

So I'm very proud to see that. We have planned to have three keynotes.

So actually we are down to two keynotes now because one of our keynote speakers is sick.

But we were planning to have Professor Laura Schreiber here from the University Clinic Würzburg,

Professor Moritz Zeiss from University Clinic Erlangen,

and also Leonardo Impert who is from Bibliotheca Herzina in Rome.

And he is actually part of the Max Planck Institute for Kunstgeschichte.

So it's really nice that we have medicine and art and from the respective fields keynote speakers.

And it's actually a really great occasion to get to know the entire work of the lab.

And if you're new and if you don't know what we are doing,

it's really a great occasion to get a good overview on what's happening here.

And what's new is that we are recording talks.

So we're recording the talks, the camera is right over there, and you have this form.

So we record the talk and then after the talk you fill this form.

And in this form you can of course give the title, presentation, speaker and so on.

And here you can assign whether you want to have the recording published or not.

So essentially you can say I don't want any of the recording choices,

then the recording will simply be deleted.

There's the choice that the recording can be made accessible via password,

then it's only essentially internal for the lab.

And then there's also the choice that it can be posted on the video portal with public access.

Then really everybody can see it and there's even the option that we can upload it to YouTube

and then it's really visible to the rest of the world.

So I hope you make use of this option.

Of course not all of the 39 talks have to be published on YouTube or something like that.

But if you're really satisfied with your presentation, then I mean you prepare all of these talks

and it would be great to also present them to a wider audience.

But of course only if you agree and if you are satisfied.

Don't cross, I want to have it on YouTube if you don't think it was a good presentation.

And then we'll just delete it.

Only if yourself are satisfied we will actually publish it.

Okay, so I also brought a couple of slides.

And this time I don't want to show something that has been going on in this research of the lab,

but I want to show something that we did in a seminar.

And this is actually the seminar that we have Monday at 8 o'clock in the morning

actually in Collegian House together with the President.

And this seminar we were working towards gamification.

You know we have this problem that for deep learning, machine learning,

we need huge amounts of annotation and in particular medical data it's really hard to obtain those.

So we started data donors to get data donation and we have to permission for crowdsourcing.

But of course these data somehow need labels.

And the task of the seminar was to create methods for actually annotating the data.

And I just want to show what kind of ideas the students came up with.

So the first result here is ScanRacer.

This was done by Jessica Diehm, Robert Herrmann, Tobias Petlwiese, and Wei Cheng Lai.

And they designed a game where you are annotating CT images.

But you're not just annotating CT images.

You try to get the best segmentation line.

And the approach is that you wrap it into a racing game.