# New insights and a computational model for understanding induced motion revealed through novel variants of the Flying Bluebottle Illusion

## TL;DR

This paper explores a visual illusion called the Dueling Orbit Illusion and introduces a computational model to understand how surrounding motion affects perceived movement.

## Contribution

The paper introduces a novel computational model to quantify and predict the effects of contextual motion on perceived trajectories.

## Key findings

- The perceived orbit of a target is distorted by surrounding inducers with offset orbiting motion.
- A computational model successfully quantifies the magnitude of contextual influence in the illusion.
- The model can be applied to study induced motion across different observer domains.

## Abstract

The Flying Bluebottle Illusion is a compelling example of how the perceived trajectory of moving objects can be greatly influenced by other motion sources in the visual scene. In this article, we present a series of simplified variants of the Flying Bluebottle Illusion in which the true motion of a target is a circular orbit around a central point. However, when a similar but offset orbiting motion trajectory is added to a set of surrounding inducers, the perceived trajectory of the target is drastically altered in both extent and direction. In other words, the perceived orbiting motion of the target is “pulled” and distorted by the orbiting motion of the inducers. For simplicity's sake, we refer to the illusory effect revealed by these dueling orbits as the Dueling Orbit Illusion. These simplified variants lend themselves to empirical study with resultant effects that can be readily modeled. Here, we present a series of case examples for how the parameters of the stimuli may be varied to yield predictable effects, describe a straightforward computational model for quantifying the magnitude of the contextual influence, and discuss how the model may be leveraged to gain insight into the phenomenon of induced motion across a range of within and between observer domains.