Computer Vision News - August 2017

33 / 56

when you meet other people?
[ smiles] Sometimes… When I realized
that we are not immune to those first
impressions, I was a little bit
empathetic that this is our human
nature. This is how we act and behave.
This is how we are navigated by our first
impressions. How sad is this? That’s
how I feel sometimes.
Your first reaction would be to see it as
an expression of the virtues of the
human spirit or the contrary?
I would say the contrary.
[ we both laugh]
It’s not that bad. It’s just natural
because we have to make a lot quick
inferences in order for us to succeed or
just to survive in the world. Those
inferences are not accurate sometimes.
Even if they are accurate, we shouldn't
base our behavior on that. It’s hard to
overcome. It’s hard to overwrite.
Would you recommend people to
change their behavior if they see that
it brings them to the wrong
conclusions?
Yes
How would they be able to judge when
they are having the wrong reaction?
There are two sides of the coin.
Everyone want to be liked, look
trustworthy, friendly, or competent. We
all want to present the best of ourselves
to other people. On the other side,
people will judge you based on your
appearance. Sometimes we don’t want
to be judged by our appearance, but we
want to be judged by our behavior and
for the person that we really are.
What are your thoughts about
research in computer vision?
Right now in computer vision, you have
a lot of exciting opportunities to do
research in industry. I think there is a
new trend for a lot of students to do
internships during the summer or even
during school years. They apply what
they learn into a job in industry. Then
they turn their research ideas into a
product. I think this is very exciting and
admirable.
Also deep learning has a lot of
potential, especially when there's big
data. It will change a lot of regions
wherever big data is available, but it’s
not the only story. We still need to
combine a lot of inference and
traditional methods or other methods
in order to complete the story.
33 Amanda Song