🤖 AI Summary
This work addresses the limitations of existing 3D generation methods, which typically produce static, opaque meshes devoid of semantic structure and procedural control, thereby hindering interactive editing. The authors propose the first end-to-end, code-native framework for 3D asset generation that leverages large language models to directly synthesize executable Blender Python scripts from multimodal inputs—text and images—yielding glTF assets enriched with named components, hierarchical assemblies, constraints, and articulated joints. Evaluated on Nova3D-Bench, a newly introduced benchmark comprising 54 diverse tasks, the method consistently generates valid programs and high-fidelity geometric assets, satisfies over 98% of numerical constraints, achieves 100% fidelity in local edits, and successfully constructs 59 functional joints, substantially outperforming current baselines.
📝 Abstract
Current 3D generative models mostly produce a final surface: a visually strong but largely opaque mesh. Interactive 3D worlds need more than a surface. They need named parts, an assembly hierarchy, measurable constraints, local edit handles, and joints for articulation. We present Nova3D, a system that generates 3D assets as executable Blender source code; the compiled mesh, a binary glTF (GLB), is treated as the artifact, not the asset. Because the output is a program, semantic handles exist at generation time rather than being recovered afterward by segmentation or rigging. We evaluate on Nova3D-Bench, a frozen, spec-grounded benchmark of 54 items across six domains and three difficulty levels with text and image inputs, against eleven baselines in four families (mesh-native, part-structured, code-native, and CAD) plus a same-LLM ablation. Nova3D produces an executable program and a valid artifact for 54/54 items. Every asset exposes named parts organized in a parent-child assembly tree; no mesh-native, CAD, or segmentation baseline exposes either. It satisfies 51/52 prompt-stated numeric and count constraints (best baseline: 11/52), passes 14/18 blinded local edits with locality preserved in 18/18, and articulates 59 joints across 12 assets at 98.3% geometric validity, where every baseline exposes zero native joints. Its geometry is competitive: it wins the structured domains in a pairwise shape-quality tournament and is second only to the strongest mesh-native model, while conceding texture realism to baked-PBR systems. The central result is representational: code-native generation turns a generated 3D object from an opaque surface into a programmable asset that downstream systems can inspect, measure, edit, and animate.