PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space
3D reconstruction and generation are commonly tackled by separate paradigms: pixel-based regression for reconstruction, and latent diffusion for generation. Recent works attempt to unify them in latent space, but with notable drawbacks: the diffusion objective is defined on latent features rather than the underlying 3D representation, and both branches suffer from information loss introduced by latent encoding, while requiring a pretrained Variational Autoencoder (VAE) or Representation Autoencoder (RAE). In this paper, we reformulate these two tasks under a unified pixel-space diffusion parad
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
- Linked via arxiv authorSensen Gao →
PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space
- Linked via arxiv authorZhaoqing Wang →
PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space
- Linked via arxiv authorQihang Cao →
PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space
- Linked via arxiv authorDongdong Yu →
PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space
- Linked via arxiv authorChanghu Wang →
PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space
- Linked via arxiv authorJia-Wang Bian →
PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space