Description of field of research:

Affective Computing is an interdisciplinary domain that aims to understand emotions and provide the machine to express emotions. Attempts to define emotion formation have been extensively investigated through multimodal data and analyses. Literature, however, focused primarily on subjective feeling and ignored emotional evolution with complex processes and modalities. Therefore, we focus on a full Component Process Model (CPM) with five components (appraisal, motivation, physiology, expression, and feeling) that was not adequately addressed by data-driven methods. The novelty of our approach is in the data-driven framework to analyze emotions triggered by playing Virtual Reality (VR) games and collecting self-reports and bio-signals from wearable devices. The project will be conducted in collaboration with a CSE HDR student, who conducts an in-progress data collection. We aim to explore the possibility of immersive VR systems as a medium to trigger varying emotional terms and intensities assuming a full CPM and investigate the role of CPM components in predicting emotions using Machine Learning (ML). These findings can integrate into adaptive human-computer interfaces, game designs, education, and more to provide a richer user understanding and experience. The project involves: supporting data collection and developing multimodal predictive models to explore the relationship between CPM and emotions.

School

Computer Science and Engineering

Research areas

Affective Computing, Virtual Reality, Artificial Intelligence

The successful candidate will work under the supervision of Dr. Gelareh Mohammadi, A/Prof Tomasz Bednarz, and HDR student Rukshani Somarathna. Data collection conducts at the EPICentre, UNSW Art & Design (http://epicentre.matters.today/) and will provide access to the required devices.

This project involves assisting in ongoing data collection, conducting an exploratory analysis, and designing and implementing machine learning algorithms for emotion recognition. Publishing high-quality outcomes can be possible while optional.

  • Multi-Componential Analysis of Emotions Using Virtual Reality (https://dl.acm.org/doi/abs/10.1145/3489849.3489958)
  • Towards Understanding Emotional Experience in a Componential Framework (https://ieeexplore.ieee.org/document/8925491) 
  • Emotions are emergent processes: they require a dynamic computational architecture (https://royalsocietypublishing.org/doi/full/10.1098/rstb.2009.0141)