SIPLAB: Sensing, Interaction & Perception Lab

The SIPLAB is a research lab at ETH Zürich working on egocentric perception, computational interaction, and behavioral AI. We advance mixed reality, robotics, and human-computer inter­action toward adaptive intelli­gence and human-aware systems. We are part of the Department of Computer Science and the AI Center.

egocentric perception scalable methods for action & motion tracking
computational interaction decoding input intent for adaptive systems
behavioral AI human affect, emotions, and traits via digital biomarkers

Projects to appear

CVPR 2026 (June)

Ultra Diffusion Poser

Diffusion Motion Tracking from Sparse Inertial Sensors
IEEE ICRA 2026 (June)

Frequency-Weighted Neural Kalman Filters

ACM CHI 2026 (April)

AutoOptimization

Automating UI Optimization through Multi-Agentic Reasoning
ACM CHI 2026 (April)

Preference-Guided Prompt Optimization

Prompt Optimization for Image Generation
ACM CHI 2026 (April)

Point & Grasp

Flexible Selection of Out-of-Reach Objects Through Probabilistic Cue Integration

Affiliations and collaborations