Time^2 framework unpacks visual perception's dual temporal dynamics

A new paper from Laurent Caplette and Frédéric Gosselin challenges the assumption that visual perception is instantaneous. The researchers argue that seeing an object actually involves two distinct temporal dimensions: processing time (the hundreds of milliseconds the brain needs to interpret retinal input) and stimulus time (the duration the object must remain in view for perception to occur). These dimensions are typically conflated in the literature and rarely studied together, leaving an incomplete picture of how vision works.

To address this, the authors present Time^2, a method built on reverse correlation that measures both temporal facets simultaneously. In their arXiv preprint (2608.04218), they demonstrate that Time^2 can precisely characterize neural phenomena including rhythmic perception, predictive processing, and coarse-to-fine sampling. By capturing how information accumulates over time and how the brain processes it, the framework provides a more rigorous constraint for future vision models. This dual-measurement approach could reshape experimental design in visual neuroscience, offering a concrete tool for disentangling the temporal dynamics that underlie everyday perception.