AI's can also learn from artificial curiosity

Share:
About the research

About the research

In us, humans, curiosity is a driving force for learning. But did you know that AI robots can also benefit from curiosity, that is, 'artificial curiosity'? Computer scientist Louis Bagot (Imec - UAntwerp) explains how this works and why it is a good idea to tell AI robots to experiment and have fun. 

Robotics
Technology
Louis Bagot
imec - UAntwerpen

Louis Bagot got his first Science&Vie Kids magazine when he was about 12, and kept reading the series in increasing levels of complexity until he was 18. The magazine sparked an interest in science that motivated him to enter an engineering school in France. Around that time, the YouTube maths channel 3blue1brown triggered a major enthusiasm not only towards mathematics, but also towards the art of explaining and teaching. In parallel, the breakthrough of DeepMind's AlphaGo beating the Go World Champion with Reinforcement Learning completed the picture: he is now researching Reinforcement Learning and completing his Ph.D. to become a University Professor.

Contact Louis

https://www.uantwerpen.be/en/staff/louis-bagot/
https://www.linkedin.com/in/louis-bagot-nao

Related videos

Robin
Amsters

LED lights: The lighthouses of the future?
GPS is a wonderful technology. Unfortunately, it is of no use in large buildings, such as hospitals and airports, because concrete blocks GPS signals. But Robin Amsters (KU Leuven - FWO) is working on a solution. He explains how lightning-fast flickering LED lights will guide us and our robots indoors in the future.
Read more
Shirley
Elprama
imec
VUB

Will an exoskeleton give me superpower?

Industrial exoskeletons can support factory and construction workers in their heavy daily tasks and prevent back pain and other work-related injuries. So why exoskeletons not yet widely used in companies? That's what Shirley Elprama (Imec - VUB) is researching: she talks to companies and informs exoskeleton-designers so that they can build better exoskeletons in the future. 
Read more
Steven
Puttemans
KU Leuven

How to teach smart computers to see and understand their surroundings
Steven Puttemans (KU Leuven) trains smart computers so that they learn to see and understand their surroundings. He for example worked on a strawberry picking robot. He explains how this works in this video.
Read more
All videos