Previous Page  19 / 24 Next Page
Information
Show Menu
Previous Page 19 / 24 Next Page
Page Background

action recognition would like to do in

the long run, which is think in longer-

term relationships between video

frames. Video doesn’t consist of one

frame, two frames – that is a very

constrained environment – so I

assume that the long-term goal is to

process the whole video. Of course, in

some tasks, that is already the case.

Not yet for action localization though.

I assume that this will be the long-

term plan. Any component that

anybody would add would…or should,

or I would like it to [she laughs] lead

towards that direction. It can be

LSTMs, or end-to-end networks or

many other things, but I assume that

what matters is the general direction.

You are obviously very passionate

about this work. Is that because you

are a passionate person or because

there is something special about it?

Both! [she laughs]

What is so special for you about this

work?

My supervisor Cordelia is a person

that thinks ahead, and she transmits

this to people around her. This project

is a way of moving forward. It is one of

the first works that perform

localization considering a spatio-

temporal continuation. There were

works some years ago, excellent works

with two streams, for example, that

served the same purpose. They used

motion, which is concatenation of

consecutive frames. Again, taking into

account the motion is as an extremely

good idea. I assume that what I really

like is that part. That it is a step

forward. Baby steps, little by little.

Where were you born?

I was born in Athens in Greece, then I

studied electrical

computer

engineering in Greece.

Did you feel like science was your

passion?

If I say yes then it’s going to be too

much, but if I say no then it’s going to

be why not.

So tell us the truth!

[

both laugh

]

It always works better.

[

both laugh

]

Okay, yes.. Obviously I will tell you the

truth. I just need to think a bit more.

Which truth will you say?

[

both laugh again

]

You know, our readers are very smart.

They will know! When did you decide

to become a researcher in science?

It happened in university. I don’t know

if it was a decision, a direction, or a

desire. It was maybe a combination of

19

Friday

Vicky Kalogeiton