Thanks a lot.

So I's acutally the turn of events every time and why not.

So let's get into the video...

Ob whichever.

Hello.

Thank you.

So. I'm sorry that I don't stand up to speak when.

But I broke my toe on Sunday.

who I am to give you some background to it.

So I trained as an engineer, really as a computer vision

and machine vision and machine hearing engineer in Cambridge.

I was in the machine learning lab, which is an engineering,

and then the graphics and interaction lab

as a research assistant in computer science.

I then moved to the EPFL, where I still am.

I submitted my thesis last Friday, in fact.

So I joined as a computer scientist.

I got a fellowship as a computer scientist,

joined a computer vision lab, Sabina Sustrunc's image

and visual representation lab.

And in 2018, EPFL decided they would make a new PhD program

in the digital humanities.

And so I transferred in 2018 into this new PhD program.

And since, I've been finishing my thesis as a visiting fellow,

first at the Harvard Renaissance Studies Center,

and then at the MaxPank for Art History in Rome,

which is where I am now.

So I've somehow moved from the computer vision world

to the other world.

And actually, my thesis is much more humanities heavy,

I would say, than I expected it to be.

And what I thought I was going to talk about when we were

organizing this thing is open problems.

And I thought, well, I would actually

repeat a presentation I gave to my own lab in Lausanne,

where I just say, here's some useful problems that we

need to solve.

Please solve them.

But then I thought, actually, that's maybe not

what you want to hear.

So if I manage to click Next, there we are.

There's a little lag here.

So yeah, yeah, yeah, it's just that the internet's

a little slow because it's on Google Sheets.

All right, so what I'll actually talk about

is basically distant reading in art history

and why computer vision and data science

have a key role in art history to have

a really close collaboration, a much closer collaboration

than simply application scenarios, let's say,