Identities are most reliably held within faces (Bruce, 2017) and criminal investigations can interview witnesses and victims to construct a visual representation from memory, known as a facial composite. However, these images are inherently error-prone and naming the depicted identities can be challenging. To improve the success of composites and enhance naming rates, research is developing a promising task where learning a composite set improves the ability to identify further images (Birchenough, in prep; Erickson & Frowd, 2024).
The current research programme will contribute to the understanding of composite recognition whilst seeking one of the first manipulations of the novel procedure to develop the effect, optimise procedures and test theoretical explanations. This can guide practical application by forensic services to ‘train’ individuals such as the police, the public and potential witnesses for improved identification rates and accuracy.
Primary goals and objectives
- Developing techniques to improve the naming of facial composite images
- Further optimising techniques for enhanced composite recognition
- Understanding the mechanisms behind facial composite recognition
