Prefer a chat interface with context about you and your work?
Predicting Gaze in Egocentric Video by Learning Task-Dependent Attention Transition