Can Information About Stiffness Perception be Inferred from Action Signals Using Models?

Area of Science:

• Neuroscience
• Haptics
• Human-Computer Interaction

Background:

• Sensory feedback is crucial for perception and action control.
• The relationship between perception and action is debated, with some illusions affecting actions and others not.
• It remains unclear if perceptual information can be solely inferred from action signals.

Purpose of the Study:

• To investigate if perceptual information can be predicted from action signals alone.
• To determine if artificial tactile skin stretch illusions' effects on stiffness perception and grip force can be modeled using only action signals.
• To identify which action signals are necessary for predicting perceptual augmentation and human-like perception.

Main Methods:

• Utilized a perceptual illusion induced by artificial tactile skin stretch.
• Recorded action signals, including grip force and motion, during a stiffness discrimination task.
• Developed predictive models using recorded action signals without stimulus information.

Main Results:

• Participant's perceptual augmentation due to skin stretch was predictable from action signals to an extent.
• A trend of increased predicted perceptual augmentation correlating with real augmentation was observed.
• Grip force was essential for predicting the augmentation effect, while motion signals were needed for human-like perception prediction.

Conclusions:

• Action signals contain significant perceptual information.
• Grip force and motion signals play distinct roles in reflecting perceptual changes and enabling human-like perception.
• Findings advance understanding of perception-action links and have implications for designing haptic feedback technologies.