EE Seminar: ScopeFlow: Dynamic Scene Scoping for Optical Flow
Speaker: Aviram Bar-Haim
M.Sc. student under the supervision of Prof. Lior Wolf
Wednesday, February 5th 2020 at 15:30
Room 011, Kitot Bldg., Faculty of Engineering
ScopeFlow: Dynamic Scene Scoping for Optical Flow
Abstract
We propose to modify the common training protocols of optical flow, leading to sizable accuracy improvements without adding to the computational complexity of the training process. The improvement is based on observing the bias in sampling challenging data that exists in the current training protocol, and improving the sampling process. In addition, we find that both regularization and augmentation should decrease during the training protocol.
Using a low parameters of-the-shelf model, the method is ranked first on the MPI Sintel benchmark among all other methods, improving the best two frames method accuracy by more than 10%. The method also surpasses all similar architecture variants by more than 12% and 19.7% on the KITTI benchmarks, achieving the lowest Average End-Point Error on KITTI2012 among two-frame methods, without using extra datasets.
