Natural visually guided behavior

Mary Hayhoe

doi:10.1016/j.brainres.2025.149899

Natural visually guided behavior

2025-08-25

0

PRE

AI

M

Mary Hayhoe *

C

Constantin A. Rothkopf

DOI：10.1016/j.brainres.2025.149899

Original

Original request for help

Save

Abstract

En 中文

The ultimate goal of vision research is to understand visual function in the natural world. However, the dominant approach has been reductionist. Different visual functions such as motion sensitivity, and depth perception have been considered separately in tightly controlled contexts. Similarly visual perception is typically considered independently of cognition and action. It is unclear the extent to which results from such experiments can generalize, an issue that has become one of increasing concern. How to approach natural vision is not obvious, however, given the diversity of natural behaviours and lack of theoretical contexts. In this paper we note that natural behaviour has many common features that can lend coherence to the issue. Visually guided actions can be considered as a sequence of decisions evolving in time that entail evaluating the state of an uncertain world, and choosing actions to satisfy behavioural goals, while taking costs and motor variability into account. This provides a unifying framework for investigating natural behaviour in more controlled conditions.

Keywords:

natural vision

visual function

decision-making

visually guided actions

uncertainty

Natural visually guided behavior

Abstract

AI Summary

Journal

Organization