Neural rendering and Inverse rendering using physical inductive biases


Thu Nguyen-Phuoc

University of Bath
Wed, May 4 2022 - 10:00 am (GMT + 7)
About Speaker

Thu Nguyen-Phuoc is a final year student in Visual Computing at the Center for Digital Entertainment, University of Bath. An architecture student that went rogue, Thu is now interested in machine learning, 3D vision and computer graphics, in particular, neural rendering and inverse rendering. During her PhD, she spent many wonderful months visiting the Smart Geometry Processing Group, University College London, under the guidance of Professor Niloy Mitra. She also did her research internship at Adobe Research, Facebook FRL Research and DeepMind.

In her previous life, she did her undergraduate degree in Architecture at the University of Bath, UK and her Master’s in Computational Design and Digital Fabrication (MSc. ITECH) at the Institute for Computational Design and Construction, University of Stuttgart, Germany.


In this talk, she will introduce her recent work on combining the expressiveness of CNNs and knowledge of the physical world in the tasks of neural rendering and inverse rendering. She will present a differentiable neural renderer that learns to render 2D images from 3D shapes directly from data. She will further describe generative models which learn 3D representations with explicit controls over objects positions and poses. Surprisingly, these representations can be learned from images alone and without supervision from pose labels, multiple views, or other 3D information.

Related seminars

Trieu Trinh

Google Deepmind

AlphaGeometry: Solving IMO Geometry without Human Demonstrations
Fri, Jul 5 2024 - 10:00 am (GMT + 7)

Tat-Jun (TJ) Chin

Adelaide University

Quantum Computing in Computer Vision: A Case Study in Robust Geometric Optimisation
Fri, Jun 7 2024 - 11:00 am (GMT + 7)

Fernando De la Torre

Carnegie Mellon University

Human Sensing for AR/VR
Wed, Apr 24 2024 - 07:00 am (GMT + 7)

Anh Nguyen

Microsoft GenAI

The Revolution of Small Language Models
Fri, Mar 8 2024 - 02:30 pm (GMT + 7)