WorldSculpt decomposes video scenes into editable 3D meshes

// 1h agoOPENSOURCE RELEASE

WorldSculpt decomposes video scenes into editable 3D meshes

Developed by Alaya Lab and the University of Tokyo, WorldSculpt is an open-source framework that reconstructs video scenes into hundreds of individually editable 3D object meshes. By conditioning single-object generative priors on multi-view inputs, it hallucinates watertight geometries for occluded objects to convert static scans into interactive environments.

// ANALYSIS

Monolithic 3D representations like Gaussian splats look visually stunning, but they remain fundamentally non-interactive; modular, object-level meshification is the true missing link required to transform passive video captures into interactive gaming environments, AR/VR worlds, and robotics training simulators. By conditioning single-object canonical priors on multi-view camera poses, WorldSculpt bypasses the compute bottlenecks of full-scene 3D diffusion and resolves severe occlusion artifacts. This approach turns static 3DGS reconstructions into modular, physics-ready assets, though overall fidelity remains tied to the precision of upstream 2D and 3D perception pipelines.

// TAGS

3d-gencomputer-vision3d-gaussian-splattingmesh-generationroboticsscene-decompositionalaya-labopen-source

DISCOVERED

1h ago

2026-09-13

PUBLISHED

1h ago

2026-09-13

RELEVANCE

7/ 10

AUTHOR

AI Search

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE1h ago

Claude Code introduces plugin eval command

Anthropic's Claude Code version 2.1.269 introduces claude plugin eval, a native testing utility built directly into the CLI to automate plugin maintenance. Running the command within a plugin directory automatically suggests test cases and runs them to verify expected plugin behavior.

OPEN SOURCE1h ago

Show-Harness lets foundation VLMs control physical robots

Show-Harness is an open-source framework developed by Show Lab at the National University of Singapore that connects vision-language models directly to robotic control systems. By translating discrete semantic action units into deterministic metric motions across diverse robot platforms, it enables zero-shot manipulation with frontier VLMs and high-frequency execution on lightweight models with minimal fine-tuning.

RESEARCH1h ago

Fire3D reconstructs simulation-ready 3D scenes from images

Fire3D is an open-source feed-forward framework that reconstructs complete, simulation-ready 3D indoor scenes from an unsegmented RGB image or casual video in under 60 seconds without per-scene test-time optimization. Instead of outputting view-dependent rendering fields, Fire3D decomposes scenes into discrete objects and background structures, generating 6-DoF poses, watertight geometry, and PBR textures tailored for robotics simulations and game engines.