Time-Agnostic Prediction: Predicting Predictable Video Frames
This addresses the challenge of visual prediction in robotics by exploiting natural predictable events, though it is incremental in applying time-agnostic methods to specific tasks.
The paper tackles the problem of predicting future or intermediate video frames by decoupling prediction from rigid time intervals, focusing instead on predictable 'bottleneck' frames. It demonstrates higher visual quality and coherent semantic subgoals in robotic manipulation tasks.
Prediction is arguably one of the most basic functions of an intelligent system. In general, the problem of predicting events in the future or between two waypoints is exceedingly difficult. However, most phenomena naturally pass through relatively predictable bottlenecks---while we cannot predict the precise trajectory of a robot arm between being at rest and holding an object up, we can be certain that it must have picked the object up. To exploit this, we decouple visual prediction from a rigid notion of time. While conventional approaches predict frames at regularly spaced temporal intervals, our time-agnostic predictors (TAP) are not tied to specific times so that they may instead discover predictable "bottleneck" frames no matter when they occur. We evaluate our approach for future and intermediate frame prediction across three robotic manipulation tasks. Our predictions are not only of higher visual quality, but also correspond to coherent semantic subgoals in temporally extended tasks.