Audiovisual detection at different intensities and delays

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Psychophysics

Background:

• Redundant signals tasks show faster responses when multiple stimuli are presented.
• Race and coactivation models, like the Wiener diffusion superposition model, attempt to explain these redundancy gains.
• Previous models struggled to accurately describe detection rates, particularly for low-intensity stimuli.

Purpose of the Study:

• To evaluate and improve diffusion superposition models for explaining redundant signal effects.
• To investigate the impact of temporal deadlines and multiple response barriers on signal detection.
• To determine which model best describes behavioral data from a simple detection task.

Main Methods:

• Two monkeys performed a simple detection task using auditory, visual, and audiovisual stimuli.
• The Wiener diffusion superposition model was modified by adding a temporal deadline (A) and a second "nogo" barrier (B).
• Closed-form and numerical solutions were derived for the modified diffusion processes.

Main Results:

• The basic Wiener diffusion superposition model poorly described the observed data, especially accuracy for low stimulus intensity.
• The modified model with a temporal deadline (A) provided a significantly better fit to the data.
• The two-barrier model (B) was outperformed by the deadline model.

Conclusions:

• A temporal deadline is crucial for accurately modeling evidence accumulation in simple detection tasks.
• The deadline model offers a superior explanation for redundancy gains compared to the two-barrier approach.
• These findings refine computational models of sensory integration and decision-making.