I have started working on real-time recognition with deep learning and I read in many articles that optical flow can't work in real-time with deep neural networks, because of computational time.
I found an article using motion vectors instead of optical flow, so I have started trying to understand the difference between those two and I found also a definition for the motion field.
Could someone explain what is the difference between the motion vector and motion field? Are they the same? Do they differentiate from optical flow in the same way?