WorldSculpt splits crowded 3D scenes into hundreds of editable meshes
WorldSculpt reconstructs grounded, posed multi-view imagery, requiring camera poses, instance masks and coarse 3D boxes rather than raw phone video.
By RuntimeWire Staff · Published · Updated
Primary source: Aligned News - AI Intelligence
Why it matters
Editable meshes turn convincing 3D captures into assets that game engines, simulators and robots can manipulate. WorldSculpt advances that conversion, though its structured inputs keep it well short of one-tap phone scanning.
Muyao Niu, project lead Zhixiang Wang (@lightchaserx), and Alaya Lab director Kaipeng Zhang released WorldSculpt, a research system that rebuilds crowded multi-view scenes as collections of separate 3D meshes instead of one fused representation.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.