Research map / Video / Video generation
LiDARCrafter: Dynamic 4D World Modeling from LiDAR Sequences
Alan Liang, Youquan Liu, Yu Yang, Dongyue Lu, Linfeng Li, Lingdong Kong, Huaici Zhao, Wei Tsang Ooi
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 dense 2D foreground mask projected from 3D boxes (ablation) | LiDAR range diffusion · nuScenes | FPD (scene): 33.97 -> 33.21 | Improved |
| add object foreground generation branch (object masks) (ablation) | LiDAR range diffusion · nuScenes | FPD (scene): 33.21 -> 24.02 | Improved |
| add sparse object conditioning (position embedding) (ablation) | LiDARCrafter · nuScenes | FPD (scene): 24.02 -> 15.97 | Improved |
| add class and box embeddings to sparse conditioning (ablation) | LiDARCrafter · nuScenes | FPD (scene): 15.97 -> 8.64 | Improved |
| add box embedding on top of position and class (ablation) | LiDARCrafter · nuScenes | FRD: 193.27 -> 194.37 | No clear effect |
| use autoregressive warped generation instead of end-to-end (ablation) | LiDARCrafter Scene2Seq · nuScenes | TTCE / CTC: TTCE(3) 3.21 -> 3.31; CTC(3) 5.68 -> 4.31 | Mixed |
| add depth and intensity history priors to autoregressive generation (ablation) | LiDARCrafter Scene2Seq · nuScenes | FRD: 311.27 -> 254.39 | Improved |
| add depth-only history prior (drop intensity) (ablation) | LiDARCrafter Scene2Seq · nuScenes | FRD: 254.39 -> 194.37 | Improved |
| drop depth prior, keep intensity (ablation) | LiDARCrafter Scene2Seq · nuScenes | FRD: 254.39 -> 364.27 | Hurt |
| use LiDARCrafter versus prior LiDAR generators on scene fidelity | LiDARCrafter · nuScenes | FRD: 243.35 (R2DM) -> 194.37 | Improved |
| use LiDARCrafter versus OpenDWM-DiT on temporal consistency (CTC) | LiDARCrafter · nuScenes | CTC: interval 1: 0.89 -> 1.12; interval 3: 3.06 -> 3.02 | Mixed |
Closest papers by meaning
- Genesis: Multimodal Driving Scene Generation with Spatio-Temporal and Cross-Modal Consistency (NeurIPS 2025 · similarity 0.89)
- OLiDM: Object-aware LiDAR Diffusion Models for Autonomous Driving (AAAI 2025 · similarity 0.89)
- LiDAR-RT: Gaussian-based Ray Tracing for Dynamic LiDAR Re-simulation (CVPR 2025 · similarity 0.88)
- Towards Realistic Scene Generation with LiDAR Diffusion Models (CVPR 2024 · similarity 0.88)
- StreetCrafter: Street View Synthesis with Controllable Video Diffusion Models (CVPR 2025 · similarity 0.88)
- Outdoor Scene Extrapolation with Hierarchical Generative Cellular Automata (CVPR 2024 · similarity 0.88)
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.
