Modeling Turn-Taking with Semantically Informed Gestures
This work addresses turn-taking prediction for conversational AI systems, but it is incremental as it extends existing datasets and methods.
The paper tackled the problem of modeling turn-taking in conversation by incorporating semantically informed gestures, showing that using a Mixture-of-Experts framework with text, audio, and gestures yields consistent performance gains over baselines.
In conversation, humans use multimodal cues, such as speech, gestures, and gaze, to manage turn-taking. While linguistic and acoustic features are informative, gestures provide complementary cues for modeling these transitions. To study this, we introduce DnD Gesture++, an extension of the multi-party DnD Gesture corpus enriched with 2,663 semantic gesture annotations spanning iconic, metaphoric, deictic, and discourse types. Using this dataset, we model turn-taking prediction through a Mixture-of-Experts framework integrating text, audio, and gestures. Experiments show that incorporating semantically guided gestures yields consistent performance gains over baselines, demonstrating their complementary role in multimodal turn-taking.