Cross-subject aesthetic preference recognition of Chinese dance posture using EEG

Area of Science:

• Neuroscience
• Art and Aesthetics
• Computer Science

Background:

• Dance aesthetic cognition is subjective due to individual differences.
• Objective criteria for dance aesthetic preference are lacking.
• Understanding the neural basis of aesthetic preference is crucial.

Purpose of the Study:

• To explore the neural mechanisms of aesthetic preference in Chinese dance postures.
• To develop a cross-subject model for recognizing aesthetic preferences in dance.
• To establish an objective measure for dance aesthetic preference.

Main Methods:

• Designed Chinese dance posture materials using Dai nationality dance.
• Collected electroencephalography (EEG) signals from 91 subjects.
• Utilized transfer learning and convolutional neural networks for EEG signal analysis.

Main Results:

• Achieved 79.74% accuracy in recognizing aesthetic preferences from EEG signals.
• Demonstrated the feasibility of an objective aesthetic measurement model for dance.
• An ablation study validated the importance of specific brain regions and hemispheres.

Conclusions:

• The occipital and frontal lobes are key areas for visual aesthetic processing of dance postures.
• The right brain plays a more significant role in processing the aesthetics of Chinese dance.
• The developed model provides a feasible objective measurement for dance aesthetic preference.