- Titel
- Functional organization of behaviorally-relevant object information in human visual cortex: Linking behavior-derived object dimensions to broadly sampled object responses in fMRI
- Zitierfähige Url:
- https://nbn-resolving.org/urn:nbn:de:bsz:15-qucosa2-1011788
- Datum der Einreichung
- 06.08.2025
- Datum der Verteidigung
- 27.11.2025
- Abstract (EN)
- How does the human visual system enable us to understand our visual world and act upon it in meaningful ways? While traditional frameworks emphasize object recognition and categorization as central computational goals guiding visual cortical organization, these approaches face basic limitations. Other perspectives propose that the visual cortex constructs richer representations of object properties relevant to diverse behavioral contexts. One approach towards behaviorally-relevant object representations is through perceived similarity judgements. However, data linking brain responses and behavior has been limited. Thus, a comprehensive account of visual cortical organization based on behavior-derived object information is yet to be established. This dissertation addresses this gap, examining how behaviorally-relevant object information captured by perceived object similarity is represented in the human brain. Study 1 introduces the THINGS-data collection, a multimodal dataset integrating extensive neuroimaging with large-scale behavioral similarity judgments of broadly sampled objects. Results demonstrated the value of this collection for object vision research and linking object representations across brain and behavior. Building on this, Study 2 links fMRI responses to thousands of object images to 66 core dimensions derived from large-scale behavioral responses of object similarity. Results revealed distributed representations of behaviorally-relevant information throughout visual cortex, capturing much of its functional architecture. The results also characterize category-selectivity as a special case in a more general multidimensional representation with more predictive power. Together, these studies advance a behavior-focused account for understanding visual cortical organization that aligns with our ability to understand visual objects.
- Forschungsdatenverweis
- THINGS-data: A multimodal collection of large-scale datasets for investigating object representations in brain and behavior
DOI: https://doi.org/10.25452/figshare.plus.c.6161151
Link: https://doi.org/10.25452/figshare.plus.c.6161151 - Freie Schlagwörter (EN)
- Object vision, visual cortex, perceived similarity
- Klassifikation (DDC)
- 570
- Den akademischen Grad verleihende / prüfende Institution
- Universität Leipzig, Leipzig
- Version / Begutachtungsstatus
- publizierte Version / Verlagsversion
- URN Qucosa
- urn:nbn:de:bsz:15-qucosa2-1011788
- Veröffentlichungsdatum Qucosa
- 18.12.2025
- Dokumenttyp
- Dissertation
- Sprache des Dokumentes
- Englisch
- Lizenz / Rechtehinweis
CC BY 4.0