Results 1 comments of VanUST

Has anyone managed to inference this model? I got pretty strange output flow map on all videos. Seems like you need to have special input image shape, but i couldn't...