Speaker Listener

Speaker-listener interaction research focuses on understanding and modeling the dynamic interplay between speakers and listeners, aiming to improve human-computer interaction and enhance our understanding of communication. Current research emphasizes developing computational models, often employing neural networks and diffusion models, to generate realistic listener responses (e.g., head movements, backchannel cues) based on speaker input, improving the realism and controllability of virtual agents. This work has significant implications for creating more engaging and natural virtual assistants, improving human-computer communication, and furthering our understanding of the cognitive processes involved in conversation.