Reclaiming saliency: Rhythmic precision-modulated action and perception
Computational models of visual attention in artificial intelligence and robotics have been inspired by the concept of a saliency map. These models account for the mutual information between the (current) visual information and its estimated causes. However, they fail to consider the circular causali...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2022-07-01
|
Series: | Frontiers in Neurorobotics |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fnbot.2022.896229/full |
_version_ | 1811311744189988864 |
---|---|
author | Ajith Anil Meera Filip Novicky Thomas Parr Karl Friston Pablo Lanillos Noor Sajid |
author_facet | Ajith Anil Meera Filip Novicky Thomas Parr Karl Friston Pablo Lanillos Noor Sajid |
author_sort | Ajith Anil Meera |
collection | DOAJ |
description | Computational models of visual attention in artificial intelligence and robotics have been inspired by the concept of a saliency map. These models account for the mutual information between the (current) visual information and its estimated causes. However, they fail to consider the circular causality between perception and action. In other words, they do not consider where to sample next, given current beliefs. Here, we reclaim salience as an active inference process that relies on two basic principles: uncertainty minimization and rhythmic scheduling. For this, we make a distinction between attention and salience. Briefly, we associate attention with precision control, i.e., the confidence with which beliefs can be updated given sampled sensory data, and salience with uncertainty minimization that underwrites the selection of future sensory data. Using this, we propose a new account of attention based on rhythmic precision-modulation and discuss its potential in robotics, providing numerical experiments that showcase its advantages for state and noise estimation, system identification and action selection for informative path planning. |
first_indexed | 2024-04-13T10:24:12Z |
format | Article |
id | doaj.art-b50440c8691a4e7c92ea91ac8a12431d |
institution | Directory Open Access Journal |
issn | 1662-5218 |
language | English |
last_indexed | 2024-04-13T10:24:12Z |
publishDate | 2022-07-01 |
publisher | Frontiers Media S.A. |
record_format | Article |
series | Frontiers in Neurorobotics |
spelling | doaj.art-b50440c8691a4e7c92ea91ac8a12431d2022-12-22T02:50:23ZengFrontiers Media S.A.Frontiers in Neurorobotics1662-52182022-07-011610.3389/fnbot.2022.896229896229Reclaiming saliency: Rhythmic precision-modulated action and perceptionAjith Anil Meera0Filip Novicky1Thomas Parr2Karl Friston3Pablo Lanillos4Noor Sajid5Department of Cognitive Robotics, Faculty of Mechanical, Maritime and Materials Engineering, Delft University of Technology, Delft, NetherlandsDepartment of Neurophysiology, Donders Institute for Brain Cognition and Behavior, Radboud University, Nijmegen, NetherlandsWellcome Centre for Human Neuroimaging, University College London, London, United KingdomWellcome Centre for Human Neuroimaging, University College London, London, United KingdomDepartment of Artificial Intelligence, Donders Institute for Brain Cognition and Behavior, Radboud University, Nijmegen, NetherlandsWellcome Centre for Human Neuroimaging, University College London, London, United KingdomComputational models of visual attention in artificial intelligence and robotics have been inspired by the concept of a saliency map. These models account for the mutual information between the (current) visual information and its estimated causes. However, they fail to consider the circular causality between perception and action. In other words, they do not consider where to sample next, given current beliefs. Here, we reclaim salience as an active inference process that relies on two basic principles: uncertainty minimization and rhythmic scheduling. For this, we make a distinction between attention and salience. Briefly, we associate attention with precision control, i.e., the confidence with which beliefs can be updated given sampled sensory data, and salience with uncertainty minimization that underwrites the selection of future sensory data. Using this, we propose a new account of attention based on rhythmic precision-modulation and discuss its potential in robotics, providing numerical experiments that showcase its advantages for state and noise estimation, system identification and action selection for informative path planning.https://www.frontiersin.org/articles/10.3389/fnbot.2022.896229/fullattentionsaliencyfree-energy principleactive inferenceprecisionbrain-inspired robotics |
spellingShingle | Ajith Anil Meera Filip Novicky Thomas Parr Karl Friston Pablo Lanillos Noor Sajid Reclaiming saliency: Rhythmic precision-modulated action and perception Frontiers in Neurorobotics attention saliency free-energy principle active inference precision brain-inspired robotics |
title | Reclaiming saliency: Rhythmic precision-modulated action and perception |
title_full | Reclaiming saliency: Rhythmic precision-modulated action and perception |
title_fullStr | Reclaiming saliency: Rhythmic precision-modulated action and perception |
title_full_unstemmed | Reclaiming saliency: Rhythmic precision-modulated action and perception |
title_short | Reclaiming saliency: Rhythmic precision-modulated action and perception |
title_sort | reclaiming saliency rhythmic precision modulated action and perception |
topic | attention saliency free-energy principle active inference precision brain-inspired robotics |
url | https://www.frontiersin.org/articles/10.3389/fnbot.2022.896229/full |
work_keys_str_mv | AT ajithanilmeera reclaimingsaliencyrhythmicprecisionmodulatedactionandperception AT filipnovicky reclaimingsaliencyrhythmicprecisionmodulatedactionandperception AT thomasparr reclaimingsaliencyrhythmicprecisionmodulatedactionandperception AT karlfriston reclaimingsaliencyrhythmicprecisionmodulatedactionandperception AT pablolanillos reclaimingsaliencyrhythmicprecisionmodulatedactionandperception AT noorsajid reclaimingsaliencyrhythmicprecisionmodulatedactionandperception |