Precision-based causal inference modulates audiovisual temporal recalibration.

Area of Science:

• Neuroscience
• Perception Science
• Computational Neuroscience

Background:

• Stable temporal perception is crucial for interacting with dynamic environments.
• Cross-modal temporal recalibration mechanisms, how the brain adjusts timing across senses, are not fully understood.

Purpose of the Study:

• To investigate the underlying mechanisms of cross-modal temporal recalibration.
• To determine how the brain updates temporal biases based on audiovisual stimulus asynchrony.

Main Methods:

• Experimentally manipulated audiovisual temporal delays presented to participants.
• Compared six computational observer models differing in bias updating and timing precision assumptions.
• Utilized Bayesian model comparison to evaluate model fit to observed recalibration data.

Main Results:

• Recalibration effects showed nonlinearity (plateauing with increasing asynchrony) and asymmetry (depending on which sense led).
• A causal-inference observer model, assuming modality-specific precision in arrival latency, best explained the observed nonlinear and asymmetric recalibration patterns.
• Findings indicate that the brain infers causality to adjust temporal perception.

Conclusions:

• Cross-modal temporal recalibration relies on causal-inference processes, similar to those in general cross-modal perception.
• Modality-specific precision of arrival latency is critical for accurate temporal recalibration.
• This research provides a computational framework for understanding audiovisual temporal perception.