Intelligent systems are pervasive in modern day life, from the apparent mind reading of the Internet to hearing aids listening in on our conversations. Understanding how man and machine interacts and how advanced data analysis can affect our productivity and personal quality of life, is the basis of our research.
At this section we research information processing in man and computer with a particular focus on the signals exchanged between the two: Audio, imagery, behaviour, and the opportunities these signals offer for modelling and prediction.
Our research is based on statistical machine learning and signal processing, on quantitative analysis of digital media and text, on mobility and complex networks, and on cognitive psychology.
Neuroinformatics research concerns new methods for analysis of neuroscience data and efforts to integrate and curate neuroscience databases.
Our neuroinformatics research has made basic contributions to neuroimaging, including the first “mind reading” methods for fMRI and PET imaging. We have introduced machine learning in neuroimaging, developed the Brede neuroimaging database, and contributed tools for resampling based optimization of neuroimaging pipelines (NPAIRS).
Current research concerns real-time methods for fMRI and EEG, connectivity modeling, and new search engine paradigms.
Lars Kai Hansen Professor, head of section
Machine learning turns abstract data into active knowledge by identifying predictive relations. Machine learning has become a major driver of the knowledge-based society.
It drives the Google economy, empowers bioinformatics, and enables mind reading in neuroimaging through neuroinformatics. Our research in machine learning is rooted in statistics and in resampling-based methods. It has a strong algorithmic component.
Current research concerns sparse representations, infinite models, multiway methods and complex networks.
Both human brains and computers are information processing systems. Many computing problems are trivial for humans while very difficult for machines, e.g. face recognition, language and content based search. Other computing problems are trivial for machines on the other hand but hard for humans, e.g. reasoning, judgment based on probabilities and handling large amounts of information.
Experimentation and modeling of cognitive psychology makes the difference between humans and machines clearer and helps us understand and improve human-computer interactions.
Currently, in the Cognitive Science and Technology lab, we study human cognition and perception through behavioral and neurophysiological experiments combined with mathematical modeling. Our purpose is to understand human cognition in a way that can be used to develop artificial cognitive systems.
Human-computer interaction (HCI) concerns design, implementation, and evaluation of systems and processes involving humans and computers. This becomes increasingly important as human well-being and productivity rely more and more on information processing and services.
Our research in HCI concerns the interaction between humans and intelligent systems with a specific focus on the signals they exchange. This includes audio, video, and physiological signals.
Our vision is to design profound cognitive systems for augmented human cognition in real-life environments. We have contributed real-life demonstrations of augmented cognition in the context of mobile applications, audio search engines, and scientific discovery.
