Manipulating Two-Dimensional Animations by Dynamical Distance Geometry

Abstract

The dynamical Distance Geometry Problem (dynDGP) was recently introduced to tackle the problem of manipulating existing animations by modifying and/or adding ad-hoc distance constraints in a distance-based representation of the motion. Although the general problem is NP-hard, satisfactory results have been obtained for the dynDGP by employing local optimization methods, where the original animations, the ones to be manipulated, are given as starting points. New animations are presented in this short paper and, differently from previous publications where only artificial instances were considered, one new animation is extracted from a video clip, depicting animated geometrical objects, that was previously used in a psychological study. The manipulation by distance constraints of such an animation allows to modify the perception of the “actions” performed by the objects of the initial animation.