Virtual Avatar

Virtual avatar research focuses on creating realistic and controllable digital representations of humans for applications in virtual and augmented reality, entertainment, and training. Current efforts concentrate on improving avatar fidelity through advanced 3D modeling techniques, such as neural radiance fields (NeRFs) and Gaussian-based methods, and enabling realistic animation from limited input data like monocular videos or sparse sensor readings. These advancements are driven by the need for more immersive and interactive experiences, impacting fields ranging from social human-computer interaction to robotics and healthcare through improved simulation and training capabilities.