Abstract
When temporally integrating information, humans are often, but not always, biased to overweight early evidence (Nienborg et al 2009, Kiani et al 2008). In a previous study (Lange et al 2018) we could explain these apparently conflicting results by assuming that the brain performs approximate inference in a hierarchical model in which expectations influence sensory inferences. Here, we use this framework to ask two related questions for which our model makes testable predictions: (1) Does perceptual decision-making adapt to the rate at which new visual information is presented, or are its assumptions about the temporal input correlations fixed, learnt over long times? (2) Does the strength of feedback differ between fovea and periphery, as recently proposed (Zhaoping et al 2017 and 2018)?
In our experiments, we show ten visual frames of band-pass-filtered noise with orientation power centered on -45 or +45 degrees, respectively. For Experiment 1, the stimulus was presented as an annulus spanning 2.1 degrees around the fixation point and we compared two variations of this experiment: one where each evidence frame had a duration of 42ms, and the other where the duration was 166ms. For Experiment 2 we kept duration of each evidence frame fixed (83ms) and varied the eccentricity of the stimulus, comparing 2.1 degrees and 9.0 degrees.
We found that the strength of the temporal integration bias differed when measured in physical time, but stayed the same in “frame-time”, indicating that the brain had adapted to the rate at which it received independent visual information. Second, we compared the strength of the primacy effect near the fovea and in the periphery, and did not find a significant difference, suggesting similar strength of feedback near the fovea and at moderately peripheral locations.