Measuring human experience
What a person judged and how that judgement formed are different questions, and self-report answers only the first. We combine psychophysics and subjective testing with EEG, EMG, eye tracking, pupillometry and autonomic measures to reach both.
Modelling human experience
Every interface assumes something about what its user will notice and how much they can take in at once. We make those assumptions explicit, as computational models of perception that a system can be designed and evaluated against.
Behaviour, intention and mental state
A system that assists someone has to know what they are doing and what they are about to do, from a signal that is always partial. We infer action, intention and emotion from video, physiology and context, in operating rooms, de-escalation training and motor rehabilitation.
Acting on machine inference
Proxy metrics and trust questionnaires do not establish whether a person understood an inference or can tell when it is likely to be wrong. We measure understanding directly, and build representations a person can act on and argue with.
Adaptive systems
Estimating a user's state while a system is running is a different problem from measuring it in the laboratory: a single pass, no trigger, a person who moves. We are building estimation that holds up under those conditions.
Human-centered design
Modality and timing are usually settled by convention and by the hardware at hand. We design and compare interaction for situations that leave little room — hands occupied, sterile field, limited connectivity, no technical training.
Participate in our studies!
We are always looking for people that are interested in taking part in our experimental studies. Our experiments mainly aim at understanding perceptual and cognitive aspects of technical systems. We use conventional psychophysical methods such as rating scales and investigate novel physiological methods such as electroencephalography (EEG), or combinations thereof.
Find more information here.