FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks скачать в хорошем качестве

FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

Optical flow estimation -- demos from supplementary material Full paper here: https://lmb.informatik.uni-freiburg.d... Complete source code here: https://github.com/lmb-freiburg/flownet2 From the paper abstract: In this paper, we advance the concept of end-to-end learning of optical flow. The large improvements in quality and speed are caused by three major contributions: 1. We focus on the training data and show that the schedule of presenting data during training is very important. 2. We develop a stacked architecture that includes warping of the second image with intermediate optical flow. 3. We elaborate on small displacements by introducing a subnetwork specializing on small motions.