Hello
given a RGB image:
image = torchvision.flow_to_image(optical_flow)
How can I reconstruct optical_flow using image alone? (in other words to create image_to_flow)
Thanks!
Hello
given a RGB image:
image = torchvision.flow_to_image(optical_flow)
How can I reconstruct optical_flow using image alone? (in other words to create image_to_flow)
Thanks!