Models
Human behaviour, physiology and health-related states
Models that interpret video, speech, language, movement, and physiological signals together.
Teodora Vuković
Building AI that understands the whole human.
I am a scientist, technology leader, and founder working at the intersection of multimodal AI and human science. I build models, datasets, and research platforms that connect video, speech, language, movement, physiology, and subjective experience.
Multimodal AI
Human Behaviour & Experience
Research Infrastructure
Scientist-Founder
My flagship venture
MOSAIC
Multimodal AI models, data, and technology for understanding human behaviour, physiology, and experience.
About
I work across computational linguistics, computer vision, speech technology, physiology, behavioural science, and research infrastructure. My work brings these perspectives together through shared models, interoperable technology, and interdisciplinary research programmes.
At the University of Zurich, I lead multimodal technology research. Alongside my academic work, I am developing MOSAIC to translate advances in multimodal AI into technology for research, healthcare, and industry.
I combine scientific leadership with product development, partnership building, grant acquisition, and the coordination of interdisciplinary technical teams.
Flagship venture
MOSAIC
Multimodal AI for human understanding
People speak, move, react and experience simultaneously, yet these signals are usually analysed separately. MOSAIC brings them together.
Platform
An AI scientist for multimodal human data
Synchronise recordings, automate annotation, run integrated analyses, and produce reproducible results.
Dataset
Behaviour, brain, body and subjective experience
A pioneering dataset connecting observable signals with people’s own accounts of what they experienced.
MOSAIC began with support from the Innosuisse Innovation Booster on Artificial Intelligence.
Research & Technology
From scientific ideas to shared infrastructure.
Venture Building
MOSAIC
Models · Platform · Dataset
Multimodal AI & Infrastructure
DAVA Lab
Building a no-code environment for automatic analysis of speech, language, facial behaviour, gaze, and movement.
Interoperability
MAVA Cluster
Connecting audiovisual research tools through shared formats and APIs so that annotations and data can move between systems.
Human Identity & Responsible AI
CAPIRE
Explainable machine learning for identity patterns in speech, facial expression, gesture, and conversational dynamics.
Open Science & Education
LaDaD
Teaching data analytics and computational methods for language data across nine Swiss institutions.
Corpus Infrastructure & Low-Resource Languages
Earlier Work
LiRI Corpus Platform, TraCeBa, the Spoken Torlak Dialect Corpus, and corpus-based research on linguistic variation.
Selected publications
Research across modalities and disciplines.
My research spans multimodal person understanding, audiovisual and speech analysis, research infrastructure, corpus linguistics, and multimodal AI.
2026
Micro-Expression-Aware Avatar Fingerprinting via Inter-Frame Feature Differencing
Proceedings of FG2026, IEEE Biometrics
2025
Beyond Appearance: Transformer-based Person Identification from Conversational Dynamics 2nd Best Paper
15th International Conference on Computer and Knowledge Engineering (ICCKE 2025), IEEE
2025
LiRI Corpus Platform: Demonstration of a Web-Based Infrastructure for Multimodal Corpus Analysis
Interspeech 2025
2026
Towards Language-Independent Face-Voice Association with Multimodal Foundation Models
FAME 2026 Challenge, ICASSP 2026
2026
TidyVoice 2026 Challenge Evaluation Plan
Interspeech 2026
Consciousness & contemplative science
I am deeply curious about human experience.
My interest in human understanding extends to consciousness, contemplative science, and first-person experience. I teach meditation and collaborate on research connecting subjective reports with language, behaviour, and neurophysiological signals.
Practice & Teaching
meditative.dev ↗
Co-founder & Meditation Teacher
First-Person Research
Meditative Phenomenology
Language analysis of first-person experience reports from advanced retreats.
Multimodal Research
Brain, Speech & Meditation
Collaborations connecting EEG, speech, biofeedback, and contemplative practice.
Work With Me
Interested in collaboration or consulting? Let’s connect.