Affective Computing for Human Wellbeing

AI-driven Human-Computer Interaction systems

Affective Computing for Human Wellbeing

The Affective Computing for Human Wellbeing area investigates the development of AI-driven Human-Computer Interaction systems. These systems are designed to elicit specific human responses, monitor these responses, and model them using statistical and machine learning techniques. The goal is to deepen our understanding and enhance human wellbeing.

  • Intelligent systems: We research the computational foundations to enable computers with the abilities to analyze, recognize, and simulate human communicative dynamics during social interactions with AI-based virtual humans. 
  • Wellbeing assessment: We create multimodal machine learning models to recognize emotions, psychological traits, mental disorders, neurodevelopment disorders, and others, using physiological (EEG, HRV, EDA…) and behavioural (voice, eye-tracking, kinetic, facial…) biomarkers.
  • Behaviour toolkits & dataset: We develop toolkits and datasets to characterize cognitive-emotional states based on signal processing and artificial intelligent techniques.



Affective Computing

Health Behaviour Informatics

Intelligent HCI Sytems

Multimodal machine learning

Technological research oriented toward engineering disciplines has developed tools, equipment and procedures that have helped to overcome some limitations of traditional approaches and even have provided solutions to specific problems.

Most of our studies involve the application of new technologies, such as (but not only) virtual reality, to old problems in neurorehabilitation and basic neuroscience. However, we have put special efforts on examining the characteristics of the technological solutions and how it affects to the users’ performance, and, although it is not our goal, we have also developed technology-driven solutions to overcome specific problems that we have faced in our research.

Signal processing

After a severe injury to the brain, individuals may present difficulties to maintain awareness of themselves and of the environment, and to respond to environmental stimuli. Recovery of consciousness usually moves from the comatose phase to an unresponsive wakefulness syndrome, where wakefulness is preserved but there are no signs of awareness, to a minimally conscious state, where subjects can present certain level of awareness although not consistent enough to enable communication, or to a locked-in syndrome, where awareness is present but no motor response.

Disorders of consciousness represent a neurologic challenge from a diagnostic, prognostic, and therapeutic point of view. Misdiagnosis reaches up to 15% to 43% of the cases, prognosis is uncertain, and no treatment has been empirically shown to be effective. During the last two decades, the increased survival and the prolonged life expectancy of these individuals have generated an increase in case incidence and prevalence, and consequently a renewed clinical and scientific concern. However, more efforts are still needed to improve current clinical knowledge about these states.

Recent Projects