To Fall Or Not To Fall: A Visual Approach to Physical Stability Prediction

TL;DR

This paper proposes a data-driven, end-to-end visual approach to predict the physical stability of wooden block towers, bypassing explicit simulation, and evaluates its performance against human judgments.

Contribution

It introduces a novel learning-based method that predicts stability directly from appearance, contrasting traditional model-based simulation approaches.

Findings

The approach accurately predicts tower stability on synthetic data.

It correlates well with human judgments on the same stimuli.

The method demonstrates potential for understanding physical reasoning visually.

Abstract

Understanding physical phenomena is a key competence that enables humans and animals to act and interact under uncertain perception in previously unseen environments containing novel object and their configurations. Developmental psychology has shown that such skills are acquired by infants from observations at a very early stage. In this paper, we contrast a more traditional approach of taking a model-based route with explicit 3D representations and physical simulation by an end-to-end approach that directly predicts stability and related quantities from appearance. We ask the question if and to what extent and quality such a skill can directly be acquired in a data-driven way bypassing the need for an explicit simulation. We present a learning-based approach based on simulated data that predicts stability of towers comprised of wooden blocks under different conditions and quantities related to the potential fall of the towers. The evaluation is carried out on synthetic data and compared to human judgments on the same stimuli.