Neural Representation and Rendering of Real-world Scene

Speaker: Dr. Lingjie Liu
         Max Planck Institute for Informatics

Title:   "Neural Representation and Rendering of Real-world Scene"

Date:    Monday, 14 February 2022

Time:    4:00pm - 5:00pm (HKT)

Zoom link:
https://hkust.zoom.us/j/928308079?pwd=MW9wTCtlSDd2MnViZGdNd2oreUpXZz09

Meeting ID:     928 308 079
Passcode:       20212022

Abstract:

High-quality reconstruction and photo-realistic rendering of real-world
scenes are two important tasks that have a wide range of applications in
AR/VR, movie production, games, and robotics. These tasks are challenging
because real-world scenes contain complex phenomena such as occlusion,
thin structures, changing topology, and biological motion. Approaching
these tasks using classical computer graphics techniques is a highly
complex and time-consuming process, requiring complicated capture
procedures and sophisticated global illumination rendering process. In
this talk, I will introduce our recent work that combines the traditional
graphics pipeline with deep learning techniques for modeling humans and
scenes in a lightweight way. Specifically, I will talk about creating
photorealistic animatable human characters from only RGB videos, and
high-quality reconstruction and novel view synthesis of general static
scenes from 2D image inputs. Finally, I will discuss challenges and
opportunities in reconstruction and rendering for future works.

*****************
Biography:

Lingjie Liu is Lise Meitner Postdoctoral Research Fellow working with
Prof. Christian Theobalt in the Visual Computing and AI Department at the
Max Planck Institute for Informatics. She received her Ph.D. degree from
the University of Hong Kong in 2019. Before that, she obtained her B.Sc.
degree in Computer Science from Huazhong University of Science and
Technology in 2014. Her research interests include neural scene
representations, neural rendering, human performance modeling and capture,
and 3D reconstruction. Webpage: https://lingjie0206.github.io/
[lingjie0206.github.io].