Generative Neural Scene Representationsfor 3D-Aware Image Synthesis

This talk covers the following 3 papers:
GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis
www.cvlibs.net/publications/Sc...
github.com/autonomousvision/graf
GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields
www.cvlibs.net/publications/Ni...
m-niemeyer.github.io/project-...
CAMPARI: Camera-Aware Decomposed Generative Neural Radiance Fields
arxiv.org/abs/2103.17269
Slides:
www.cvlibs.net/publications/Ni...

[@lesleyHsieh]

Fantastic research! The presentation is very clear.
Thank you very much for the sharing!

[@yttan757]

Very helpful presentation!
So the "learned camera pose" in your third work actually solves the limitation presented in Giraffe, that model tends to entangle dataset bias when doing object rotation?
The portrait no longer face the camera like Mona Lisa does.

[@gc-uk7jb]

Geiger, my Superman !