Unsupervised Intuitive Physics from Past Experiences
This addresses the challenge of enabling AI systems to quickly adapt physical reasoning from limited experiences, which is incremental but relevant for robotics and navigation.
The paper tackles the problem of learning intuitive physics models from videos in an unsupervised, meta-learning framework, achieving generalization across time, space, and varying numbers of objects.
We are interested in learning models of intuitive physics similar to the ones that animals use for navigation, manipulation and planning. In addition to learning general physical principles, however, we are also interested in learning ``on the fly'', from a few experiences, physical properties specific to new environments. We do all this in an unsupervised manner, using a meta-learning formulation where the goal is to predict videos containing demonstrations of physical phenomena, such as objects moving and colliding with a complex background. We introduce the idea of summarizing past experiences in a very compact manner, in our case using dynamic images, and show that this can be used to solve the problem well and efficiently. Empirically, we show via extensive experiments and ablation studies, that our model learns to perform physical predictions that generalize well in time and space, as well as to a variable number of interacting physical objects.