An EEG-based framework for exploring adaptive rhythmic human-machine interaction

Area of Science:

• Cognitive Science
• Human-Computer Interaction
• Neuroscience

Background:

• Existing experimental paradigms for rhythmic interaction often lack ecological validity and holistic analysis.
• Limitations include unrealistic partner behavior, inflexible design, and insufficient user experience analysis, hindering insights into human-human rhythm dynamics.

Purpose of the Study:

• To present and validate a novel multimodal paradigm for evaluating human-human rhythm interaction.
• To extend this paradigm for controlled evaluation of interactions with virtual AI agents.
• To address limitations of existing methods by improving ecological validity and partner realism.

Main Methods:

• Participants engaged in a tapping paradigm with audio-visual drum animations.
• Partners were either human or AI-driven, under simple and complex (polyrhythmic) conditions.
• Portable electroencephalography (EEG) and post-trial questionnaires assessed neural and subjective responses.

Main Results:

• The paradigm demonstrated improved ecological validity compared to existing methods.
• Partner identity (human vs. AI) was effectively masked, maintaining positive user experiences (flow, arousal, enjoyment).
• Portable EEG successfully measured neural modulation and temporal alignment, supporting unobtrusive assessment.

Conclusions:

• The validated paradigm provides a flexible foundation for studying rhythmic interaction in human-machine systems.
• It balances ecological realism with experimental control, paving the way for adaptive and biofeedback systems.
• The AI-driven drummer represents a significant first step for future virtual rhythm interaction research.