Summary
- Major: Neuroscience
- Faculty Mentor: Dr. Fang Jiang
- Research Topic: Investigating the Other Race Effect Using Racially Transformed Faces
Abstract
The other-race effect (ORE), a well-documented psychological phenomenon where people are generally better at recognizing and remembering faces of their own race compared to those of other races, has
been widely demonstrated using real face images with both recognition memory and perceptual matching paradigms, with no prior research employing artificially generated faces. This project will examine the ORE using artificially generated faces as a way to validate our RaceGAN framework, which converts racial appearance in face images while preserving the face identity. Demonstrating the ORE with these synthetic faces will help assess the framework’s effectiveness and explore how racially translated images can be used to evaluate and reduce bias in face recognition and facial attribute classification systems. The ability of the visual system to adapt to facial characteristics plays an important role in how we recognize and interpret others. Although critical, achieving appropriate levels of adaptation to faces from a particular race often requires extensive experience and may take a lifetime to fully develop. The current study introduces a novel approach in which the images, rather than the observers, are adapted and tests whether modifying visual stimuli to approximate optimal states of adaptation can enhance face perception performance.