CUDA-GHR: Controllable Unsupervised Domain Adaptation for Gaze and Head Redirection.

The robustness of gaze and head pose estimation models is highly dependent on the amount of labeled data. Recently, generative modeling has shown excellent results in generating photo-realistic images, which can alleviate the need for labeled data. However, adopting such generative models to new domains while maintaining their ability to provide fine-grained control over different image attributes, e.g., gaze and head pose directions, has been a challenging problem. This paper proposes CUDA-GHR, an unsupervised domain adaptation framework that enables fine-grained control over gaze and head pose directions while preserving the appearance-related factors of the person. Our framework simultaneously learns to adapt to new domains and disentangle image attributes such as appearance, gaze direction, and head orientation by utilizing a label-rich source domain and an unlabeled target domain. Extensive experiments on the benchmarking datasets show that the proposed method can outperform state-of-the-art techniques on both quantitative and qualitative evaluations. Furthermore, we show that the generated image-label pairs in the target domain effectively transfer knowledge and boost the downstream tasks' performance.