Visual Cognition Research Lab

On October 28th, Megan Lawrence successfully defended her PhD Thesis, entitled "How do the Socio-Cognitive Impacts of Real vs. AI-Generated Facial Expressions of Emotion Differ?” Congratulations Dr. Lawrence!

Dr. Corina Lacombe

Dr. Megan Lawrence, B.A., M.A., Ph.D.

Doctoral Student, Experimental Psychology
E-mail: mlebl128@uottawa.ca

Megan very recently defended her PhD thesis!  She began her work as a PhD student in the lab in September 2020. Before that she was an honours student under the supervision of our frequent collaborator, Isabelle Boutet. Megan's research program concerns how AI-driven image generators, such as Stable Diffusion, generate face images of people of colour. Her findings show that these programs reflect biases prevalent in society, generating more realistic images of white individuals and reflecting stereotypes such as the "angry black woman" and the "submissive asian person". 

Sarah began her PhD work here in 2022. Her research interests concern the effects of schizophrenia and bipolar disorder on facial expression processing. In pursuing her work, Sarah has used a range of techniques, including meta-analysis, EEG measurements, and behavioural testing.

Chiyo is doing her honours thesis work on the effects of various psychopathology symptoms on facial expression processing.

Khadidja is doing her honours thesis work on the effects of gender on emotion perception in biological motion.

Robert is doing his honours thesis work on the perception of facial expressions of pain.

Led by doctoral student Sarah MacNeil, this meta-analysis explores whether there are deficits in the accuracy and reaction time of people with disorders on the schizophrenia spectrum when they engage in emotional expression recognition tasks. It is well known that people with schizophrenia are worse at recognizing the emotions of others, however, less is known about the emotion recognition abilities of people with other disorders on the schizophrenia spectrum (i.e., schizoaffective disorder, psychotic disorder). Although disorders on the schizophrenia spectrum are similar in their symptoms, they have key diagnostic criteria that differentiate them, so it is important to understand whether changes in social functioning arise from such differences. The findings of this project will provide a better understanding of the kinds of things people on the schizophrenia spectrum have difficulty with, which in turn will help us to develop treatments to improve social functioning across the schizophrenia spectrum. 

This collaboration with colleague Isabelle Boutet examines the degree to which Emoji can function as a stand-in for non-verbal communication in face-to-face interactions. When people interact in person, their spoken words are supplemented by a wide range of information from their facial expressions, tone of voice, speech cadence, and body language, among other things. In text-based interactions, such as via SMS, Twitter, and other social media, these cues are lacking. It has been suggested that an over-reliance on these low bandwidth forms of communication may harm emotional intelligence and social connectivity, leading to isolation and anti-social behaviour. It is thought that emoji may be able to compensate here, filling the gap left by non-verbal cues, but very few studies to date have examined how effective they are. In this series of studies, which is only just beginning, we aim to find out to what degree emoji are successful in transmitting socially important information that is normally carried by non-verbal cues in face-to-face interactions. 

Led by doctoral student Corina Lacombe, this project examines whether social anxiety disorder (SAD) is associated with a negative cognitive and attentional bias to various emotional facial expressions. Some research suggests that ambiguous facial cues, such as neutral emotional expressions, are perceived negatively by people with anxiety. Furthermore, there is evidence to suggest that these perceptual biases contribute to the maintenance of the disorder. However, inconsistent findings in the literature have led Corina to conduct a meta-analysis investigating whether individuals with SAD are able to correctly identify and categorize various emotional facial expressions. The findings of this project will help us better understand some of the maintaining factors associated with social anxiety disorder.

It has been suggested that humans engage in micro-expressions, which are subtle short-duration expressions of emotions such as anger, joy and fear. It has been further suggested that people can be trained to read these subtle cues and thus gain insight into the true feelings of others by seeing through fake smiles or forced neutral expressions.  But little work has examined whether the human visual system is even fast enough to pick up these cues. Led by doctoral candidate Justin Chamberland, this project examines the speed with which individuals can read briefly-presented emotional expressions.  Our results show that humans can detect certain emotional expressions, such as happiness, in as little as 5 milliseconds.  This opens the door to the possibility that micro-expressions can indeed be read. However, to date our stimuli have been high-intensity full-face expressions, not the subtler partial ones that are typical of real-world micro-expressions. So more work remains to me be done.  Stay tuned...

The emotional enhancement of memory (EEM) effect refers to the phenomenon by which people tend to remember emotionally-charged stimuli better than emotionally neutral ones. In this project, we examine the extent to which the EEM effect applies to facial recognition. We are testing this using a diverse sample of face stimuli from across the age spectrum, courtesy of the FACES Database. Additionally, using eye-tracker technology (EyeLink 1000), we aim to uncover how eye movements correlate with facial recognition. The present study is part of a larger ongoing research project that will examine age effects by testing elderly participants. However, initial testing will be conducted on a sample composed of young people, who have been shown in the literature to display a negativity bias when exposed to facial stimuli portraying emotional expressions. Therefore, we expect angry faces to be better remembered than either happy or neutral ones among this population. Past research in facial recognition has tended to focus on either facial identity or emotional expression processing. The present study aims to provide a significant contribution to existing research by combining both elements in order to examine their interaction.

Non-Suicidal Self-Injury, commonly known as "cutting", refers to the intentional self-inflicted damaging of one's body without the intent to die. It is a pathological behaviour with an increasing and high prevalence among teens and young adults. This behaviour appears under the category of “conditions for further study” within the DSM-5, which encourages further research into the condition. Within her current research, doctoral candidate Laura Ziebell examines whether this condition is associated with differences in recognizing and correctly categorizing facial expressions of emotion, which some theories of its etiology suggest should be the case. Our findings (Ziebell et al., 2018) show that people with a history of NSSI are better able to recognize facial expressions, in that they can correctly categorize more subtle expressions. This is despite the fact that they seem to have less of a tendency to reflexively mimic others' positive and negative emotions, which is known to be helpful in emotion recognition and can promote social relationships (Ziebell et al., in press). The findings of this study help us better understand a common and troubling behaviour that disproportionately affects young people, and which is a risk factor for a number of negative outcomes, including suicide. Ultimately, it may lead to an increased understanding of this condition and improvements in therapeutic interventions used to target this population.