In this talk, I will discuss our work related to design of generative models, which allow for realistic generation of talking heads. We have placed emphasis on disentangling motion from appearance and have learned motion representations directly from RGB, without structural representations such as facial landmarks or 3D meshes. We have aimed at constructing motion as linear displacement of codes in the latent space. Based on this, our model LIA (Latent Image Animator) and LIA-X are able to animate images via navigation in the latent space, allowing for control over generation.

While highly intriguing, video generation has thrusted upon us the imminent danger of deepfakes, which can offer unprecedented levels of increasingly realistic manipulated videos. Deepfakes pose an imminent security threat to us all, and to date, deepfakes are able to mislead face recognition systems, as well as humans. Hence, we design generation and detection methods in parallel.