1: \begin{abstract}

2: 

3: % Recent research has demonstrated automated facial analysis tasks to often perform poorly for inter-sectional racial groups due to the presence of racial bias, which gives significance to the analysis and characterisation of human facial diversity. The existing literature presents 3D modelling and generation of facial images as an effective means of control over changing facial attributes. However, 

4: 

5: 

6: Achieving an effective fine-grained appearance variation over 2D facial images, whilst preserving facial identity, is a challenging task due to the high complexity and entanglement of common 2D facial feature encoding spaces. Despite these challenges, such fine-grained control, by way of disentanglement is a crucial enabler for data-driven racial bias mitigation strategies across multiple automated facial analysis tasks, as it allows to analyse, characterise and synthesise human facial diversity. In this paper, we propose a novel GAN framework to enable fine-grained control over individual race-related phenotype attributes of the facial images. Our framework factors the latent (feature) space into elements that correspond to race-related facial phenotype representations, thereby separating phenotype aspects (e.g. skin, hair colour, nose, eye, mouth shapes), which are notoriously difficult to annotate robustly in real-world facial data. Concurrently, we also introduce a high quality augmented, diverse 2D face image dataset drawn from CelebA-HQ for GAN training. Unlike prior work, our framework only relies upon 2D imagery and related parameters to achieve state-of-the-art individual control over race-related phenotype attributes with improved photo-realistic output. 

7: 

8: % This pathway to disentangled facial feature representation in turn results in improved interpretability, generalisation and faster model convergence on such facial analysis tasks. 

9: 

10: \end{abstract}

11: