# From visual snapshots to allocentric representations of a 3D world > **NIH NIH F32** · UNIVERSITY OF PENNSYLVANIA · 2024 · $74,284 ## Abstract PROJECT SUMMARY/ABSTRACT Spatial neuroscience has uncovered a great deal about how animals—primarily rodents—form allocentric (world-centered) spatial maps of the world. Rodents explore by moving their bodies through the world. In contrast, primates explore the world visually, and recent work suggests this could dramatically impact their formation of allocentric maps. The current project investigates the brain systems that allow humans to form allocentric maps of 3D spaces from afar, without physically visiting the locations being represented. This work will use fMRI to detect map-like representations of viewed locations in a 3-dimensional “scene space” and to compare these to map-like representations of visited locations in a virtual “navigable space”. This work is divided into 2 aims, the first optimized for experimental power and the second optimized for ecological validity. Aim 1 is to identify map-like representations of scene space and understand their relationship to cognitive maps of navigable space. Representational similarity analysis (RSA) will be used to quantify map-like allocentric representations of locations in scene space. Experiment 1a will identify brain regions that support viewpoint-independent (allocentric) maps of viewed locations in scene space. Experiment 1b will use an analogous procedure to identify brain regions that support viewpoint-independent maps of navigable space and relate them to viewpoint-independent maps of scene space. Aim 2 is to identify representations of scene space and navigable space using dynamic, ecologically relevant tasks. This aim will leverage voxelwise encoding models to detect map-like neural representations of viewed and visited locations while participants actively explore their environments. Experiment 2a will look for maps of scene-space by scanning participants while they view the virtual courtyard from different viewpoints and actively searching for a hidden target item. Experiment 2b will look for maps of navigable space by scanning participants while they freely navigate through the virtual environment searching for hidden rewards. This research will take place in the ideal training environment for the applicant. The sponsorship team's expertise in the field of spatial navigation, advanced fMRI analysis and designs, and the computational modeling of freely- navigating participants data are critical to the applicant's training goals. Further training is available through coursework in machine learning and data science, computing resources, and a diverse set of relevant lab meetings and seminars. This work addresses a critical gap in the literature between the fields of vision science and spatial neuroscience. Understanding the role of the visual system in forming spatial maps in sighted individuals could inform interventions aimed at mitigating navigational and other challenges faced by those with low vision or other differences in sensory processing. ## Key facts - **NIH application ID:** 10901351 - **Project number:** 1F32EY036266-01 - **Recipient organization:** UNIVERSITY OF PENNSYLVANIA - **Principal Investigator:** Anna Shafer-Skelton - **Activity code:** F32 (R01, R21, SBIR, etc.) - **Funding institute:** NIH - **Fiscal year:** 2024 - **Award amount:** $74,284 - **Award type:** 1 - **Project period:** 2024-08-01 → 2027-07-31 ## Primary source NIH RePORTER: https://reporter.nih.gov/project-details/10901351 ## Citation > US National Institutes of Health, RePORTER application 10901351, From visual snapshots to allocentric representations of a 3D world (1F32EY036266-01). Retrieved via AI Analytics 2026-05-21 from https://api.ai-analytics.org/grant/nih/10901351. Licensed CC0. --- *[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*