Philip Häusser wanted to find a solution to the problem himself. “I was immediately hooked,” he recalls. “The potential that lies in what is known as Deep Learning, or Artificial Intelligence, is huge.” Using online learning programs, he initially tried to acquire know-how on Machine Learning and neural networks through self-study. However, he soon discovered that he could not get such complex knowledge properly through tutorials, which were still rare at the time and sometimes difficult to access. An internship with experts was what was needed. Philip Häusser presented his work to Prof. Dr. Daniel Cremers, who is heading the Department of Computer Vision & Artificial Intelligence at TUM. He did not take him on as an intern. Instead, he offered him a doctoral position.

BREAKING NEW GROUND

Starting in 2014, Philip Häusser and his doctoral adviser conducted research on artificial neural networks in the field of Computer Vision. With FlowNet, they were among the first researchers to figure out how to use Deep Learning for Optical Flow Estimation in an image sequence. Optical flow is an important representation of motion information in the early stages of image processing and is used, for example, in the visual navigation of robots or autonomous driving. “In life and in science, you sometimes have to venture into the unknown to get ahead,” says Philip Häusser. “At TUM, I met a lot of people who have broken new ground. That made a lasting impression on me and remains a very significant inspiration for me to this day.”

Philip Häusser did not pass up the opportunity to be part of a small team and drive this innovative idea forward from the very beginning. He is now Chief Technology Officer and Managing Director of the start-up, which is growing steadily with a staff of fifty and successful Series B financing rounds.

KNOWLEDGE FOR EVERYBODY

Alongside his interest for Artificial Intelligence, Philip Häusser has another great passion. Ever since he was a schoolboy, he has been involved as a TV presenter, book author and producer of web videos to bring science to the general public in a comprehensible and entertaining way. “In Germany, we conduct cutting-edge research that has a global impact, but we talk about it way too little,” he explains. “In some parts of society, there is occasionally downright distrust of science, as we saw increasingly during the Corona pandemic.”

Whether for ZDF, BR Fernsehen or on YouTube channels such as Terra X Lesch & Co and Faktencheck, Philip Häusser is presenting scientific content and astounding experiments with great enthusiasm. In 2021, he published his second book of popular science. “I feel it’s my duty as a scientist to explain how all this actually works – so that knowledge isn’t exclusive to a few people, but everyone can join in the conversation,” he says. “That kind of empowerment is what drives and motivates me.”