Reasoner
What the Missing Frames Showed: Machine Learning Describes Masked Video Events
Neural networks can describe in words what’s happening in pictures and videos — but can they make sensible guesses about things that happened before or will happen afterward? Researchers probed this ability.