Using hearing and vision for motion prediction, motion perception, and localization

Area of Science:

• Cognitive Psychology
• Neuroscience
• Human Perception

Background:

• Accurate localization of moving objects is crucial for daily activities, particularly in traffic.
• Understanding how humans integrate multisensory information (auditory, visual) for object localization, especially under challenging conditions like occlusion, is not well understood.

Purpose of the Study:

• To investigate how humans utilize auditory and visual information to localize moving and static objects.
• To determine if and how multisensory integration for localization is affected by expected sensory interference, such as occlusion.
• To compare human performance in audiovisual target localization with predictions from maximum likelihood estimation (MLE).

Main Methods:

• Four experiments were conducted involving auditory, visual, and audiovisual targets.
• Participants localized targets under conditions with audiovisual and visual-only occluders.
• Localization performance was assessed and compared to maximum likelihood estimation (MLE) predictions.

Main Results:

• When an audiovisual occluder was present, participants relied solely on visual cues, ignoring auditory information.
• With a visual-only occluder, participants exclusively used auditory cues, disregarding visual information.
• While localization estimates aligned with MLE predictions, no precision benefits were observed for moving audiovisual targets; however, static audiovisual targets showed near-MLE integration.

Conclusions:

• Humans employ unisensory input for localizing moving objects and predicting motion during occlusion, flexibly prioritizing senses based on anticipated modality-specific interference.
• Multisensory integration, showing significant benefits, is utilized for static object localization, aligning with MLE models.
• This adaptive sensory prioritization enhances object localization in dynamic and potentially noisy environments.