Human Understanding and Empathy: emotion color wheel
Return to Microsoft Research Lab – Redmond

HUE: Human Understanding and Empathy

The mission of the HUE team is to empower people by inventing new technologies that promote emotional resilience and well-being. We achieve this by developing ideas and technologies that lead to systems having a richer understanding of people’s knowledge, capabilities and context, and thus fueling intelligent systems and experiences that improve people’s lives.

Emotions are fundamental to human interaction and they influence learning, memory, decision-making, well-being, and other aspects of our lives. Computing can be a powerful tool in enabling people to be more productive, healthier, emotionally resilient. By using intelligent systems that leverage behavioral, contextual and physiological signals, we can build emotionally responsible systems. We are an interdisciplinary team working on different areas including HCI, Machine Learning, Psychology, and we are passionate about bringing emotional intelligence to technology.

Workplace stress in our modern society has been quickly increasing during the last few decades which has only worsened by the unique demands imposed by COVID and remote/hybrid work settings. These conditions are detrimental to the mental health and wellness of workers and lead to huge business costs in terms of productivity and medical costs around the world. To help alleviate this problem, the HUE team is advancing research around the following topics:

  • Passive Sensing. Creating new sensing methods to comfortably capture relevant information about the user’s stress (e.g., respiration, heart rate variability).
  • Contextualized Understanding. Developing effective methods to aggregate multimodal user and contextual data in a personalized and meaningful manner to understand stress.
  • Personalized Interaction. Evaluating different types of potential support considering the originating source of stress, work habits, and users’ preferences.
  • Human-AI Collaboration. Designing meaningful interactions between the sensing, analytics, and recommendation systems and the users whose data the systems are reasoning about.
  • Affective Computing Ethics. Defining appropriate uses of AI in the context of Affective Computing to help support users while ensuring their maximum privacy and respect.