Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval
This work addresses the problem of precise facial manipulation and retrieval for computer vision applications, representing a significant but incremental advance over existing StyleGAN-based methods.
The paper tackles unsupervised facial feature transfer and retrieval on real images by introducing Retrieve in Style (RIS), which improves feature disentanglement for challenging transfers like hair and pose, eliminates per-image tuning, and enables fine-grained face retrieval, achieving high-fidelity results in analyses.
We present Retrieve in Style (RIS), an unsupervised framework for facial feature transfer and retrieval on real images. Recent work shows capabilities of transferring local facial features by capitalizing on the disentanglement property of the StyleGAN latent space. RIS improves existing art on the following: 1) Introducing more effective feature disentanglement to allow for challenging transfers (ie, hair, pose) that were not shown possible in SoTA methods. 2) Eliminating the need for per-image hyperparameter tuning, and for computing a catalog over a large batch of images. 3) Enabling fine-grained face retrieval using disentangled facial features (eg, eyes). To our best knowledge, this is the first work to retrieve face images at this fine level. 4) Demonstrating robust, natural editing on real images. Our qualitative and quantitative analyses show RIS achieves both high-fidelity feature transfers and accurate fine-grained retrievals on real images. We also discuss the responsible applications of RIS.