Rhea Bergman

[Flea @flea333]

I am very sad that the West African Black Rhinoceros has gone extinct.

[Alex Bergman @alexwbergman]

For more info, reach out to me (here or at NeurIPS) or check out our webpage: https://t.co/BX2NztOR8G
Special thanks to collaborators: Petr Kellnhofer, @toomanyyifans, @ericryanchan, @DaveLindell, @GordonWetzstein
#NeurIPS22 #NeurIPS (7/7)

[Alex Bergman @alexwbergman]

In the paper, we show that this same framework also works for faces, enabling the generation of editable radiance field representations of faces. (6/7)

[Alex Bergman @alexwbergman]

The explicit separation of these two sources of variance allows 3D GANs to perform better on more complex datasets, like bodies. Additionally, the resulting generated 3D models can be deformed using the same deformation process used at training! (5/7)

[Alex Bergman @alexwbergman]

The pose-conditioned deformation module allows us to render images of the body identity in any potential pose. As long as the distribution of poses matches that of the ground truth dataset, the generator can learn a latent space modeling only the identities of humans. (4/7)

[Alex Bergman @alexwbergman]

In practice, the generator generates a radiance field representation of a body in a canonical pose conditioned on the identity vector. However, when rendering this radiance field, a pose-conditioned deformation module is used to determine where rays sample in the volume. (3/7)

[Alex Bergman @alexwbergman]

To do this, we factorize our neural radiance field generator architecture to independently model body identity and pose. This means that our generator takes in two inputs: one from a latent space which learns to model body identities, and one which models body poses. (2/7)

[Alex Bergman @alexwbergman]

We are excited to present our work at #NeurIPS22 on Generative Neural Articulated Radiance Fields! In this paper, we build on top of the capabilities of existing 3D GAN architectures and propose the first method for generating full human bodies! (1/7)

[Flea @flea333]

Imagine being a little scurrying winter mouse in a Finnish village

[Rhea Bergman @rheabergman]

INCREDIBLE 😤😤😤😩😩

[Alex Bergman @alexwbergman]

Check out our new paper "Fast Training of Neural Lumigraph Representations using Meta-learning":
https://t.co/u6Fpek5Fsx

We propose MetaNLR++, which is able to train and render neural scene representations in a fraction of the time that competing methods require!

[Flea @flea333]

I just want art

[Flea @flea333]

Fucking shit up with black coffee, beans, and potatoes

[Bleacher Report @BleacherReport]

NBA PLAYERS OF THE WEEK 🗣

Ingram: 26.7 PTS | 7.3 REB | 6.7 AST

Sabonis: 24.3 PTS | 11 REB | 7 AST

[Friedrice Nietzsche @tinynietzsche]

AP: dow jones falls 400 points after hearing of the death of rocker eddie van halen

[kb @ojvhteuhyhes]

me fighting my way through a depression spiral in the smoking area

[On This Day In LotR @onthisdayinLotR]

Gandalf reaches the Shire.

[Flea @flea333]

The voice of god.

[Flea @flea333]

I wish my middle name was Motherfucking

[Flea @flea333]

@Cotton_216 @thejunebugfreak No echo and the bunnymen?

[Flea @flea333]

No matter what, page’s verse riff on heartbreaker makes it all worth it