Research map / Video / Video generation
InfiniCube: Unbounded and Controllable Dynamic 3D Driving Scene Generation with World-Guided Video Models
Yifan Lu, Xuanchi Ren, Jiawei Yang, Tianchang Shen, Zhangjie Wu, Jun Gao, Yue Wang, Siheng Chen, Mike Chen, Sanja Fidler, Jiahui Huang
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 pixel branch to the voxel-only reconstruction (dual-branch lifting) | InfiniCube reconstruction · Waymo novel view synthesis (T+5) | PSNR: 19.90 -> 20.80 | Improved |
| use 3D grounded guidance buffers for long video generation instead of HD-map projection conditioning | SVD-XT video model · Waymo long video (FID by frame index) | FID | Improved |
| use guidance buffers rather than Panacea HD-map projection for map alignment at frame 120 | InfiniCube video · HD map alignment user study | positive rate: 53.4% -> 84.8% | Improved |
| add semantic buffer to guidance conditions (ablation) | world-guided video model · Waymo long video | FID | Improved |
| add coordinate buffer to guidance conditions (ablation) | world-guided video model · Waymo long video | FID | Improved |
| add road surface condition to voxel generation (ablation) | voxel LDM · voxel world generation | qualitative drivable-region localization | Improved |
| reuse overlapping latents during outpainting (vs naive outpainting) (ablation) | voxel outpainting · voxel world generation | chunk transition consistency | Improved |
Closest papers by meaning
- SEM-ROVER: Semantic Voxel-Guided Diffusion for Large-Scale Driving Scene Generation (ECCV 2026 · similarity 0.90)
- X-Scene: Large-Scale Driving Scene Generation with High Fidelity and Flexible Controllability (NeurIPS 2025 · similarity 0.89)
- WorldSplat: Gaussian-Centric Feed-Forward 4D Scene Generation for Autonomous Driving (ICLR 2026 · similarity 0.89)
- Generative Rendering: Controllable 4D-Guided Video Generation with 2D Diffusion Models (CVPR 2024 · similarity 0.88)
- GEN3C: 3D-Informed World-Consistent Video Generation with Precise Camera Control (CVPR 2025 · similarity 0.87)
- Architect: Generating Vivid and Interactive 3D Scenes with Hierarchical 2D Inpainting (NeurIPS 2024 · 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.
