Michael Niemeyer: Generative Neural Scene Representations | 3D Representation Seminar

Michael Niemeyer is a Ph.D. student at the Max Planck Institute, supervised by Andreas Geiger.
His research focuses on continuous representations and generative models. In particular, He is interested in how we can infer 3D representations from sparse observations.
PAPER LINKS
GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis: arxiv.org/abs/...
GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fields: arxiv.org/abs/...
SEMINAR INFO:
Seminar website: yenchenlin.me/...

[@zhuohengli8416]

Doing presentation in my AI course about GIRAFFE, wonderful explanation, thank you for your work

[@njatonjiva5307]

Excellent explaination and discussion. Keep going

[@bookerml8502]

I hane a simple question. Dose the camera pose randomly choose from distribution? If the camera pose is random, how can we control the view when novel view synthesis？