From Graphics to Vision with Machine Learning
I am interested in rendering virtual realities (VR), how computers can understand the complex world we live in, and how these interfaces may enhance our everyday life. For this talk, I picked my favorite contributions to computer vision (CV) and computer graphics (CG) and explain how machine learning (ML) was essential to all of them. On the graphics side, I will talk about motion-driven interfaces for controlling movie and game characters with simple gestures. On the CV side, I will introduce algorithms that estimate the human motion from plain videos, recorded with lightweight egocentric and mobile cameras. Along the way, I will explain the utilized ML algorithms, from simple linear regression to advanced deep learning.
Helge Rhodin is a postdoctoral researcher at the CVLab of Pascal Fua and lecturer at EPFL. He attained his PhD at the Max Planck Institute for Informatics, advised by Christian Theobalt. He received his MSc and BSc degree in Computer science from Saarland University, where he was supervised by Joachim Weickert, Philipp Slusallek, Michael Breuß, and Sebastian Hack. He interned at Basler Vision Technologies.