Two keynotes on the generation and detection of deepfakes by Antitza Dantcheva
- Research
- Science and society
Published on December 8, 2025
–
Updated on December 8, 2025
Dates
from November 21, 2025 to November 27, 2025
Location
Sophia Antipolis, France - Sheffield, UK
In November, the 3IA Chairholder delivered two keynotes on this hot topic at two major events: SophI.A Summit 2025 (Sophia Antipolis) and SRBS@BMVC2024 (Sheffield).
Abstract:
Antitza Dantcheva talked about her work related to design of generative models, which allow for realistic generation of talking heads. They 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. They 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.
Learn more about SophI.A Summit 2025
Learn more about SRBS@BMVC2025
Antitza Dantcheva talked about her work related to design of generative models, which allow for realistic generation of talking heads. They 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. They 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.
Learn more about SophI.A Summit 2025
Learn more about SRBS@BMVC2025
