Buchumschlag

Towards Interactive Intelligence for Digital Humans

Fuente: arXiv

Gespeichert in:

Bibliographische Detailangaben
Hauptverfasser:	Cai, Yiyi, Chu, Xuangeng, Gao, Xiwei, Gong, Sitong, Huang, Yifei, Kang, Caixin, Li, Kunhang, Liu, Haiyang, Liu, Ruicong, Liu, Yun, Ng, Dianwen, Su, Zixiong, Wu, Erwin, Wu, Yuhan, Yan, Dingkun, Yan, Tianyu, Zeng, Chang, Zheng, Bo, Zhou, You
Format:	Preprint
Veröffentlicht:	2025
Schlagworte:	Computer Vision and Pattern Recognition Computation and Language Graphics Human-Computer Interaction
Online-Zugang:	Acceder al recurso
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

_version_	1866918385785241600
author	Cai, Yiyi Chu, Xuangeng Gao, Xiwei Gong, Sitong Huang, Yifei Kang, Caixin Li, Kunhang Liu, Haiyang Liu, Ruicong Liu, Yun Ng, Dianwen Su, Zixiong Wu, Erwin Wu, Yuhan Yan, Dingkun Yan, Tianyu Zeng, Chang Zheng, Bo Zhou, You
author_facet	Cai, Yiyi Chu, Xuangeng Gao, Xiwei Gong, Sitong Huang, Yifei Kang, Caixin Li, Kunhang Liu, Haiyang Liu, Ruicong Liu, Yun Ng, Dianwen Su, Zixiong Wu, Erwin Wu, Yuhan Yan, Dingkun Yan, Tianyu Zeng, Chang Zheng, Bo Zhou, You
contents	We introduce Interactive Intelligence, a novel paradigm of digital human that is capable of personality-aligned expression, adaptive interaction, and self-evolution. To realize this, we present Mio (Multimodal Interactive Omni-Avatar), an end-to-end framework composed of five specialized modules: Thinker, Talker, Face Animator, Body Animator, and Renderer. This unified architecture integrates cognitive reasoning with real-time multimodal embodiment to enable fluid, consistent interaction. Furthermore, we establish a new benchmark to rigorously evaluate the capabilities of interactive intelligence. Extensive experiments demonstrate that our framework achieves superior performance compared to state-of-the-art methods across all evaluated dimensions. Together, these contributions move digital humans beyond superficial imitation toward intelligent interaction.
format	Preprint
id	arxiv_https___arxiv_org_abs_2512_13674
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	Towards Interactive Intelligence for Digital Humans Cai, Yiyi Chu, Xuangeng Gao, Xiwei Gong, Sitong Huang, Yifei Kang, Caixin Li, Kunhang Liu, Haiyang Liu, Ruicong Liu, Yun Ng, Dianwen Su, Zixiong Wu, Erwin Wu, Yuhan Yan, Dingkun Yan, Tianyu Zeng, Chang Zheng, Bo Zhou, You Computer Vision and Pattern Recognition Computation and Language Graphics Human-Computer Interaction We introduce Interactive Intelligence, a novel paradigm of digital human that is capable of personality-aligned expression, adaptive interaction, and self-evolution. To realize this, we present Mio (Multimodal Interactive Omni-Avatar), an end-to-end framework composed of five specialized modules: Thinker, Talker, Face Animator, Body Animator, and Renderer. This unified architecture integrates cognitive reasoning with real-time multimodal embodiment to enable fluid, consistent interaction. Furthermore, we establish a new benchmark to rigorously evaluate the capabilities of interactive intelligence. Extensive experiments demonstrate that our framework achieves superior performance compared to state-of-the-art methods across all evaluated dimensions. Together, these contributions move digital humans beyond superficial imitation toward intelligent interaction.
title	Towards Interactive Intelligence for Digital Humans
topic	Computer Vision and Pattern Recognition Computation and Language Graphics Human-Computer Interaction
url	https://arxiv.org/abs/2512.13674