Research map / 3D vision / 3D generation
WonderZoom: Multi-Scale 3D World Generation
Jin Cao, Hong-Xing Yu, Jiajun Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add scale-aware opacity modulation (ablation) | WonderZoom · Scale-aware opacity modulation ablation | GPU memory: 7.96G -> 3.40G | Improved |
| add scale-aware opacity modulation (ablation) | WonderZoom · Scale-aware opacity modulation ablation | FPS: 1.4 -> 97.2 | Improved |
| add scale-aware opacity modulation (ablation) | WonderZoom · Zoom-in rendering quality | visual blur | Improved |
| add scale-consistent depth registration (ablation) | WonderZoom · New-scale depth estimation (beetle example) | shape distortion | Improved |
| add auxiliary view synthesis (ablation) | WonderZoom · Complete 3D scene from novel views | missing regions | Improved |
| use WonderZoom instead of Voyager (CLIP score) | WonderZoom · Multi-scale 3D world generation (32 scenes, 8 images) | CLIP score: 0.2609 -> 0.3432 | Improved |
| use WonderZoom instead of Gen3C (NIQE) | WonderZoom · Multi-scale 3D world generation (32 scenes, 8 images) | NIQE: 4.924 -> 3.695 | Improved |
| use WonderZoom instead of WonderWorld (generation time) | WonderZoom · Multi-scale 3D world generation (32 scenes, 8 images) | Time per new-scale scene (s): 9.3 -> 62.1 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Gen3C — GEN3C: 3D-Informed World-Consistent Video Generation with Precise Camera Control (CVPR 2025) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- WonderWorld: Interactive 3D Scene Generation from a Single Image (CVPR 2025 · similarity 0.91)
- Scale3D: Autoregressive Modeling for Large Outdoor Scene Generation (ECCV 2026 · similarity 0.89)
- WorldGen: From Text to Traversable and Interactive 3D Worlds (CVPR 2026 · similarity 0.89)
- WorldGrow: Generating Infinite 3D World (AAAI 2026 · similarity 0.89)
- SGAM: Building a Virtual 3D World through Simultaneous Generation and Mapping (NeurIPS 2022 · similarity 0.89)
- A 3D Scene is Worth 1K Tokens: 3D-Grounded Representation for Scene Generation at Scale (NeurIPS 2026 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
