PREDICTING BODY MOTION

For each of a plurality of time steps: receive a reference joint pose of an articulated entity and receiving an indication that another joint of the articulated entity is unobserved or observed. Prompt a motion model using the reference joint pose and a mask token. The model predicts body motion comprising a trajectory of the articulated entity and a pose of a plurality of joints of the articulated entity. The mask token represents the other joint and is temporally adaptable by: in response to receiving an indication that the other joint is unobserved, using information about the reference joint pose and a pose of the other joint from a previous time step; and in response to receiving an indication that the other joint is observed, using information about the reference joint pose and a pose of the other joint from the current time step.