Deblurring and 3D reconstruction of fast moving objects

Datum konání: 14.12.2022
Přednášející: Denys Rozumnyi, Ph.D. student at ETH Zurich
Odpovědná osoba:

Denys is a former bachelor and master student of Jiri Matas at CMP, CVUT. After finishing his master thesis, which received Werner von Siemens Award for the best thesis, he has moved to ETH, Zurich, for Ph.D. studies under the supervision of Marc Pollefeys. His work on deblurring fast moving objects attracted a lot of media attention, when his DeFMO neural network was used to prove that an image of alleged U.F.O. is a "only" a flying falcon. In his talk, he will explain how DeFMO works and show his recent results of 3D reconstruction of flying objects from blurred videos.