Detecting Hodo and Zu in Bunraku Nonverbal Communication using Hilbert-Huang Transform

Nonverbal communication relies on human actions to convey information through signals. In Bunraku puppetry performance, a UNESCO Intangible Cultural Heritage, three puppeteers manipulate a single puppet using the rhythmic structure of "Hodo" and "Zu". The puppeteer master embeds cues, "Hodo" and "Zu", into the puppet's movements, guiding coordination. However, the mechanics of these cues remain unclear. Despite this, shared gestures and rhythms enable seamless collaboration. This paper proposes a signal analysis framework using Principal Component Analysis and Hilbert-Huang Transform to investigate how puppeteers achieve nonverbal communication, synchronizing movements and rhythmic variations to manipulate the puppet with natural fluidity. The correlation analysis using Procrustes Analysis, Mutual Information, and Canonical Correlation Analysis, along with the examples identified through this method, all meet statistical significance across these three validation approaches, confirming the effectiveness of our method.