Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study introduces the Dyadic Distance Matrix (DDM) to effectively model conversational entrainment by capturing pairwise similarities in speaker interactions across multiple dimensions. By employing a speaker-switch test, the authors demonstrate that real DDMs can be distinguished from switched versions, confirming that the DDM encodes genuine interaction-specific structures rather than merely reflecting individual speaker traits. The findings underscore the importance of prosodic variability in natural conversations and validate the DDM as a diagnostic tool for assessing dyadic interactions.
Real conversational dynamics can be disentangled from individual speaker traits using a novel speaker-switch test, revealing the true nature of dyadic interactions.
Speakers in dialogue continuously adapt their communicative behavior across acoustic, lexical, and semantic dimensions, a phenomenon known as conversational entrainment. Modeling this process requires representations that capture the global structure of interaction, yet prior approaches fail to disentangle dyad-specific patterns from speaker-specific traits, limiting their ability to capture true conversational adaptation. We address this with the Dyadic Distance Matrix (DDM), which encodes all pairwise similarities between the turns of two speakers over an entire conversation, capturing long-range cross-speaker dependencies. This raises a key question: does the DDM represent genuine interaction, or merely reflect individual speaker characteristics? We propose the speaker-switch test, a principled control in which one speaker's turns are replaced with those from an unrelated speaker drawn from a different conversation. This preserves turn-level statistics while disrupting the original dyadic coadaptation. The ability to distinguish real from switched DDMs thus directly evaluates whether the representation encodes interaction-specific structure. Across four embedding types and classifiers including ResNet-50 on the CANDOR corpus, real DDMs are consistently distinguishable from their switched counterparts. Comparisons with LibriSpeech show higher discriminability in read speech, highlighting the role of prosodic variability in naturalistic conversations. GradCAM analysis further reveals distinct structural signatures driving classification. These results establish the speaker-switch test as a robust diagnostic for validating representations of dyadic conversational interaction.