A theory and model of the neural transformations mediating human object perception

Introduction

In the blink of an eye, our brain rapidly transforms the photons hitting our retina into a rich and detailed percept of the world as consisting of objects. By knowing what the objects are, and in what configuration these objects appear to us, we understand the meaning of the visual world around us. Yet, despite intense research, how neural transformations enable rich object perception remains unclear. The overall goal of the research program TRANSFORM is to provide an explanatory theory of the neural transformations mediating rich object perception, and a predictive quantitative model embodying this theory.

Research Goals

Towards this goal, TRANSFORM will provide three strong and novel constraints for theory and model building:

1. Neural Constraint: TRANSFORM will reveal the neural transformations underlying visual object perception in the mature brain.
2. Behavioural Constraint: It will unravel the link between the neural transformation and object-related behaviour.
3. Developmental Constraint: It will clarify the developmental trajectory of neural transformations underlying visual object perception from infancy into adulthood.

Methodology

For maximal efficiency and power in unified theory formation and model building, TRANSFORM will employ an integrated, interdisciplinary research strategy that combines:

• Large-scale non-invasive brain imaging to capture neural transformations in space and time with unprecedented depth.
• Advanced mathematical analysis to reveal the geometry of the transformations.
• Computational modelling using deep neural networks to build a predictive, quantitative model of those transformations.

Conclusion

Through this orchestrated effort, TRANSFORM will provide the empirical pieces of evidence for a new theory and model of the neural transformations mediating our rich everyday experience of object vision and change the way we think about and investigate human vision and cognition.