A computational deep learning investigation of animacy perception in the human brain
Abstract The functional organization of the human object vision pathway distinguishes between animate and inanimate objects. To understand animacy perception, we explore the case of zoomorphic objects resembling animals. While the perception of these objects as animal-like seems obvious to humans, s...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-12-01
|
| Series: | Communications Biology |
| Online Access: | https://doi.org/10.1038/s42003-024-07415-8 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559175068385280 |
|---|---|
| author | Stefanie Duyck Andrea I. Costantino Stefania Bracci Hans Op de Beeck |
| author_facet | Stefanie Duyck Andrea I. Costantino Stefania Bracci Hans Op de Beeck |
| author_sort | Stefanie Duyck |
| collection | DOAJ |
| description | Abstract The functional organization of the human object vision pathway distinguishes between animate and inanimate objects. To understand animacy perception, we explore the case of zoomorphic objects resembling animals. While the perception of these objects as animal-like seems obvious to humans, such “Animal bias” is a striking discrepancy between the human brain and deep neural networks (DNNs). We computationally investigated the potential origins of this bias. We successfully induced this bias in DNNs trained explicitly with zoomorphic objects. Alternative training schedules failed to cause an Animal bias. We considered the superordinate distinction between animate and inanimate classes, the sensitivity for faces and bodies, the bias for shape over texture, the role of ecologically valid categories, recurrent connections, and language-informed visual processing. These findings provide computational support that the Animal bias for zoomorphic objects is a unique property of human perception yet can be explained by human learning history. |
| format | Article |
| id | doaj-art-d3023d79687e4f758b1ee95f940fcbce |
| institution | Kabale University |
| issn | 2399-3642 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Communications Biology |
| spelling | doaj-art-d3023d79687e4f758b1ee95f940fcbce2025-01-05T12:43:17ZengNature PortfolioCommunications Biology2399-36422024-12-017111510.1038/s42003-024-07415-8A computational deep learning investigation of animacy perception in the human brainStefanie Duyck0Andrea I. Costantino1Stefania Bracci2Hans Op de Beeck3Brain and Cognition, Faculty of Psychology and Educational Sciences, KU LeuvenBrain and Cognition, Faculty of Psychology and Educational Sciences, KU LeuvenCenter for Mind/Brain Sciences (CIMeC), University of TrentoBrain and Cognition, Faculty of Psychology and Educational Sciences, KU LeuvenAbstract The functional organization of the human object vision pathway distinguishes between animate and inanimate objects. To understand animacy perception, we explore the case of zoomorphic objects resembling animals. While the perception of these objects as animal-like seems obvious to humans, such “Animal bias” is a striking discrepancy between the human brain and deep neural networks (DNNs). We computationally investigated the potential origins of this bias. We successfully induced this bias in DNNs trained explicitly with zoomorphic objects. Alternative training schedules failed to cause an Animal bias. We considered the superordinate distinction between animate and inanimate classes, the sensitivity for faces and bodies, the bias for shape over texture, the role of ecologically valid categories, recurrent connections, and language-informed visual processing. These findings provide computational support that the Animal bias for zoomorphic objects is a unique property of human perception yet can be explained by human learning history.https://doi.org/10.1038/s42003-024-07415-8 |
| spellingShingle | Stefanie Duyck Andrea I. Costantino Stefania Bracci Hans Op de Beeck A computational deep learning investigation of animacy perception in the human brain Communications Biology |
| title | A computational deep learning investigation of animacy perception in the human brain |
| title_full | A computational deep learning investigation of animacy perception in the human brain |
| title_fullStr | A computational deep learning investigation of animacy perception in the human brain |
| title_full_unstemmed | A computational deep learning investigation of animacy perception in the human brain |
| title_short | A computational deep learning investigation of animacy perception in the human brain |
| title_sort | computational deep learning investigation of animacy perception in the human brain |
| url | https://doi.org/10.1038/s42003-024-07415-8 |
| work_keys_str_mv | AT stefanieduyck acomputationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain AT andreaicostantino acomputationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain AT stefaniabracci acomputationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain AT hansopdebeeck acomputationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain AT stefanieduyck computationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain AT andreaicostantino computationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain AT stefaniabracci computationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain AT hansopdebeeck computationaldeeplearninginvestigationofanimacyperceptioninthehumanbrain |