Talking Head Generation
Talking Head Generation is a subtask of computer vision that focuses on generating dynamic speaking head portraits from a set of static facial images. The goal of this task is to synthesize realistic and coherent facial animations through deep learning technology, thereby achieving a natural human-computer interaction experience. Its application value lies in virtual anchors, video conferencing, and the entertainment industry, where it can significantly enhance user immersion and engagement.
100 sleep nights of 8 caregivers
Ashok
VoxCeleb1 - 1-shot learning
Few-shot Adversarial Model
VoxCeleb1 - 32-shot learning
Few-shot Adversarial Model
VoxCeleb1 - 8-shot learning
Few-shot Adversarial Model
VoxCeleb2 - 1-shot learning
Fast Bi-layer Avatars (medium size)
VoxCeleb2 - 32-shot learning
VoxCeleb2 - 8-shot learning
CainGAN