Speaker: Dr. Christina Carlisi, Postdoctoral Fellowship, University College

Time: 14:30-16:00, Nov. 19, 2019

Venue: Room 1113, Wang Kezhen Building

Abstract: One of the key ways affective information is received is through facial emotional expressions. However, most research involving the processing of facial emotions relies on the use of a fixed stimulus set of predefined standard facial emotional expressions. There are likely discreet differences in the way individuals perceive and interpret these expressions. Affective processing is also implicated in the development of a number of psychiatric difficulties. The emerging field of computational psychiatry has recognised the need for more fine-grained approaches to the way we study individual differences in affective processing. For example, it remains unexplored whether population-level differences in affective biases are due to more nuanced individual variation in emotion perception and how such individual differences relate to resilience to developing mental illness. This talk will present recent work developing a computational approach to explore individual differences in people’s characterisation of emotions represented by facial expressions. We use an ‘evolutionary algorithm’ to quickly and efficiently evolve facial emotional expressions, achieving an individualised, quantifiable representation of facial emotion for each participant. Ultimately, a better understanding of differences in how people interpret emotional information may tell us important information about why some people are more vulnerable to developing mental illness and help these individuals earlier in development.

Host: Prof. Xiaolin Zhou