POSTPONED Physiological I/O

Motivation

Personal mobile computing has witnessed immense progress since handheld PCs. Current wearable computing takes the form of smart eyewear, textile, and jewelry that have proximal and constant contact with our skin, which can sense and stimulate physiological activity directly.

How can computing systems be engineered to treat physiological activity as meaningful inputs, and could the same systems exert a meaningful influence on physiological activity in return?

Our bodies respond in certain ways when we are aroused, fatigued, attentive, disoriented, afraid, etc. Our pupil sizes, eye movements, skin conductance, heart rhythms, respiration, posture, and cortical activity change in accordance to how we perceive, think, and feel about the world that surrounds us. Recent advances in signal processing and pattern recognition allow the accurate measurement and rapid reconstruction of such activity, even outside a laboratory environment. Nonetheless, having access to these signals is merely the first step towards understanding what they mean and, potentially, influencing them directly.

The Dagstuhl Seminar for “Physiological I/O” will host experts across topics such as human-computer interaction, signal processing, machine learning, electrical and biomedical engineering, as well as cognitive science, psychology, and neuroscience. Group discussions and workshops will focus on the following challenges:

• Interpretation of physiological activity, given context
• Data quality of sensing in-the-wild
• Classification approaches for non-stationary user states
• Stimulation of physiological activity for relevant use-cases
• Consolidation of existing software libraries for robustness
• Ethics, privacy, and data management

The seminar will commence with a morning session of short 3-minute talks by the attendees. This will define the knowledge gaps, acknowledged challenges, and viable approaches that will be discussed in smaller workshops. There will also be keynotes throughout the seminar on the abovementioned challenges to set the theme for breakout workshops, which should identify research bottlenecks that can be worked on by individuals or smaller groups. Participants are encouraged to organize smaller meetings to develop solutions and recommended working practices as well as initiate joint writing sessions to draft research grant proposals, tutorials, workshops, and research papers. After lunch every day, paired attendees will go for a walk where they will discuss literature of joint interests. A social excursion will be planned for Wednesday afternoon.

We expect discussions to raise challenges with a near- to mid-term perspective, (i.e., 3–10 years) to arise in the course of our discussions, hence generating ideas for industry-academia collaborations and international research grant proposals. The long-term perspective is to derive a close-loop system schematic, which describes the physiological inputs and outputs of the human user to a computing system, that will guide the development of physiology-based human-machine interfaces across daily use-case scenarios.