Christoph Lassner, PhD Student
Bernstein Center for Computational Neuroscience (University of Tuebingen), Max Planck Institute for Intelligent Systems

  You are here: Research » That's my Science » Christoph Lassner, PhD Student
2 Christoph
1 Christoph

Christoph Lassner, PhD Student

Hometown: Augsburg, Germany

Affiliation: Bernstein Center for Computational Neuroscience (University of Tuebingen), Max Planck Institute for Intelligent Systems

What are you working on?

I am teaching computers to perceive humans in plain digital images. While you probably know face recognition from your digital camera, I want to have computers recognize the entire human body. At the same time, they should not only localize important keypoints describing the body pose (which is currently the most common approach), but find a finer description of the body. Simply put, this task is similar to finding a good 3D model for the body, like a digital avatar.

Why is it interesting?

This perception task must be solved every day in many situations. We humans excel at it, partly due to having a lot of practice, partly due to our still superior learning ability compared to today‘s computers. Both aspects make it very interesting: on the one hand, it is a task with high practical relevance (with autonomous cars being just one, currently popular, example), on the other hand we can learn a lot about perception and learning while solving it.

How are you going to find out?

Other researchers at the Perceiving Systems Department have developed a parameterized 3D model for the human body. This can be used for visualizing the body, but we can also integrate it into the optimization methods used for understanding images. Like that, a computer can 'imagine' a human body while figuring out the details of its picture.

Why did you decide to become a scientist?

I am passionately curious and very excited about our future. Technology opens up many possibilities for humanity to develop, and while pushing its limits, we are learning so much about ourselves: how perception and learning works! It is an interesting time to delve into machine learning, and that was true even before the deep learning hype (and my time as a PhD student) started.

Your favourite thing outside of science?

I must answer with two things: music and sports. Currently, I am more regularly pursuing sports, because they provide a better means of balance for my body. With the B12 climbing site in Tübingen and Beachvolleyball courts around the corner of the MPI, outdoor sports are well-covered. And Yoga helps me to stay more calm and centered.

 

 

Questions?     Claudia Däfler       presse-is@tuebingen.mpg.de
Updated 04/10/2016 9:37am