Closed juanfe0705 closed 4 years ago
Try another model. The different pretrained models in this repository perform differently depending on whether your input has small or large motions, homogeneous regions, or black backgrounds. If you tested a Sintel sample, I would wager a guess that you did not run the full FlowNet2.0 model.
(closed due to inactivity)
Everything works fine for me, I can generate .flo without trouble and I can convert them to .png However, when I see the .png image it looks very pixelated as in Figure 1 (left) of this paper:
but I would like to have the image as in figure 1 (right)
can you help me?