Machine Heart

Aug 19, 2026 · Artificial Intelligence

Geometry‑Grounded Video Diffusion for 3D‑Consistent World Generation

DreamWorld introduces a two‑stage Geometry‑then‑Appearance pipeline that first generates geometry features for the target viewpoint using a 3D foundation model and then conditions a video diffusion model on these features to produce high‑quality, 3D‑consistent RGB videos, achieving state‑of‑the‑art results on multiple benchmarks.