Virtual Audience

Virtual audience research focuses on creating realistic and engaging simulated audiences for various applications, primarily aiming to improve communication effectiveness and user experience in virtual environments. Current research explores using large language models (LLMs) to simulate audience reactions and predict communication outcomes, as well as developing systems to aggregate and synthesize individual audience feedback into a cohesive acoustic experience, often leveraging machine learning models like XGBoost for prediction and audio generation models for synthesis. This work has implications for enhancing online communication, improving performance in virtual events, and providing valuable insights into audience behavior and engagement.