Perzeptive Systeme Talk Biography
07 June 2022 at 12:00 | Virtual (Zoom)

Reconstruction and Synthesis for Dynamic Humans and Scenes

ORGANIZERS
Thumb ticker sm 24 07 15 smile front face
Perzeptive Systeme
  • Guest Scientist
Thumb ticker sm bild
Perzeptive Systeme
  • Research Scientist
Tianyeli

This thesis focuses on automated systems to capture realistic 4D visual content for general humans and scenes, such that we can animate and replay the captured content. Firstly, we design a system to reconstruct and register a large quantity of high-quality 4D faces across identities, expressions, and poses by utilizing geometric, photometric, and motion cues. Based on well-curated datasets we propose a lightweight yet expressive face model that works on a wide range of populations, by separately modeling the shape (identity), expression, and poses of human faces. Secondly, we design an inference network based on volumetric features and architecture to produce high-quality face registrations without the traditional photogrammetry and mesh registration. Thirdly, we propose a differentiable renderer that enables reasoning 3D geometry, poses, appearance and lighting from only images by modeling the non-differentiable rasterization as a continuous process. Lastly, we provide a novel and efficient method to capture challenging 4D scenes with the dynamic neural radiance field and efficient training strategies, by exploiting the spatial-temporal redundancy of the datasets.

Speaker Biography

Tianye Li (University of Southern California)

Ph.D.