Creation and testing of computer-generated faces for cognitive experiments

Study level

Vacation research experience scheme

Topic status

We're looking for students to study this topic.


Ms Lauren Fell
Associate Lecturer
Division / Faculty
Science and Engineering Faculty


The way humans perceive faces is an important factor in social systems. The shape and features of a person’s face help us predict the types of personality traits a person might have.

Artificial faces can be generated with features that prompt certain judgements such as trust, dominance and attractiveness. However, the images traditionally used in studies that investigate these effects are quite outdated and would benefit from being updated based on the current improvements in computer generated imaging.

This project provides the opportunity to create computer-generated faces and to test them with real human participants. You will have the freedom to integrate your own ideas and methods into the project and will be valued for your creative input.

Research activities

We aim to create an updated set of faces that vary on traits, such as dominance, trust and attractiveness, either by:

  • updating existing stimuli to appear more photorealistic
  • creating completely new stimuli by performing analyses on existing and new data sets.

We will then validate the created faces through the design and execution of a crowdsourced study with human participants.


It is expected that, at a minimum, a set of computer-generated faces will be created. An additional outcome is that these faces will be validated empirically.

Skills and experience

To be considered for this project, it is essential that you have:

  • some prior experience in 3D modelling or CAD
  • skills or the willingness to acquire skills in face morphing and feature manipulation
  • a passion for learning and developing new skills to solve a problem.

It is desirable if you possess:

  • Python skills for image machine learning
  • knowledge of machine learning in the application of face recognition.



Contact the supervisor for more information.