A unified diffusion transformer jointly solves single-image relighting and 3D reconstruction via a new isotropic NDC-Orthographic Depth representation and mixed synthetic/real training.
Humanolat: A large-scale dataset for full-body human re- lighting and novel-view synthesis
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
GeoRelight: Learning Joint Geometrical Relighting and Reconstruction with Flexible Multi-Modal Diffusion Transformers
A unified diffusion transformer jointly solves single-image relighting and 3D reconstruction via a new isotropic NDC-Orthographic Depth representation and mixed synthetic/real training.