High-resolution Face Swapping via Latent Semantics Disentanglement
This work addresses the challenge of producing realistic, high-resolution face swaps for applications in media and entertainment, representing an incremental improvement over prior methods.
The paper tackles the problem of entangled semantics in high-resolution face swapping by disentangling latent semantics using a pre-trained GAN's progressive layers, separating structure and appearance attributes, and achieves state-of-the-art results in hallucination quality and consistency for image and video swapping.
We present a novel high-resolution face swapping method using the inherent prior knowledge of a pre-trained GAN model. Although previous research can leverage generative priors to produce high-resolution results, their quality can suffer from the entangled semantics of the latent space. We explicitly disentangle the latent semantics by utilizing the progressive nature of the generator, deriving structure attributes from the shallow layers and appearance attributes from the deeper ones. Identity and pose information within the structure attributes are further separated by introducing a landmark-driven structure transfer latent direction. The disentangled latent code produces rich generative features that incorporate feature blending to produce a plausible swapping result. We further extend our method to video face swapping by enforcing two spatio-temporal constraints on the latent space and the image space. Extensive experiments demonstrate that the proposed method outperforms state-of-the-art image/video face swapping methods in terms of hallucination quality and consistency. Code can be found at: https://github.com/cnnlstm/FSLSD_HiRes.