I am a Senior Researcher & Lecturer at the Computer Graphics Laboratory of ETH Zurich, and a Research Consultant at Disney Research. I am leading the Digital Character AI projects at CGL. My research interests include conversational digital characters, affective computing, human-computer interaction, and applied machine learning.
My vision is to create intelligent digital humans that can naturally communicate, understand, and support people across domains such as education and mental health. My research focuses on multimodal artificial intelligence for interactive digital humans, developing models that combine large language models, affective computing, and data-driven animation to create embodied conversational agents endowed with autonomous agency, consistent values, and beliefs.
My work bridges machine learning, human–computer interaction, and computer graphics to enable AI systems such as Digital Einstein and interactive patient avatars for psychotherapy training and health education.
A system enabling users to "talk to their health data" through an embodied conversational agent using a dual-agent architecture, compared against traditional dashboards in a pilot …
A transformer pipeline for text-driven facial animation exploiting phoneme-level speech structure, achieving real-time performance and best-in-class lip synchronization accuracy. …
A new multimodal dataset and architecture combining egocentric vision and physiological signals for in-the-wild emotion and personality recognition, presented at NeurIPS 2025 …
A dynamic cognitive framework for narrative agents in interactive storytelling, combining BDI representations with LLM generation to balance story coherence with player agency. …
BEE is a modular cognitive framework for conversational agents featuring belief management, value alignment, transparent reasoning, and extensibility. Best Paper Honorable Mention …
A joint framework modeling personality and emotion for personality-consistent conversational agents, using contrastive learning to decouple emotion from semantic content. IVA 2025. …
A full-pipeline platform for interactive AI character experiences, demonstrated through Digital Einstein and deployed at scientific conferences, technology events, and public …
SIGGRAPH Asia 2024 Emerging Technologies demonstration describing the physical installation and AI integration of Digital Einstein at the Tokyo venue.
EmoSpaceTime decouples emotion and content in 3D speech animation through contrastive learning, enabling fine-grained control over emotional expressivity independent of spoken …
Systematic study of multimodal emotion recognition in natural human-chatbot interactions, evaluating text, acoustic, and behavioral signal fusion strategies. ICMI 2024.