High-fidelity digital humans are increasingly used in interactive applications, yet achieving both visual realism and real-time responsiveness remains a major challenge. We present a high-fidelity, real-time conversational digital human system that seamlessly combines a visually realistic 3D avatar, persona-driven expressive speech synthesis, and knowledge-grounded dialogue generation. To support natural and timely interaction, we introduce an asynchronous execution pipeline that coordinates multi-modal components with minimal latency. The system supports advanced features such as wake word detection, emotionally expressive prosody, and highly accurate, context-aware response generation. It leverages novel retrieval-augmented methods, including history augmentation to maintain conversational flow and intent-based routing for efficient knowledge access. Together, these components form an integrated system that enables responsive and believable digital humans, suitable for immersive applications in communication, education, and entertainment.
@article{arxiv.2511.12662,
title = {Hi-Reco: High-Fidelity Real-Time Conversational Digital Humans},
author = {Hongbin Huang and Junwei Li and Tianxin Xie and Zhuang Li and Cekai Weng and Yaodong Yang and Yue Luo and Li Liu and Jing Tang and Zhijing Shao and Zeyu Wang},
journal= {arXiv preprint arXiv:2511.12662},
year = {2025}
}
Comments
Proceedings of the Computer Graphics International 2025 (CGI'25)