Research map / 3D vision / Gaussian splatting
ReconDreamer++: Harmonizing Generative and Reconstructive Models for Driving Scene Representation
Guosheng Zhao, Xiaofeng Wang, Chaojun Ni, Zheng Zhu, Wenkang Qin, Guan Huang, Xingang Wang
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 depth loss with fused static point clouds on novel views (ablation) | ReconDreamer · Waymo, lane shift 3m | FID: 93.56 -> 77.51 | Improved |
| add depth loss with fused static point clouds on novel views (ablation) | ReconDreamer · Waymo, lane shift 0m | LPIPS: 0.1519 -> 0.1207 | Improved |
| add separate ground Gaussian model (ablation) | ReconDreamer + depth loss · Waymo, lane shift 3m | FID: 77.51 -> 75.22 | Improved |
| add separate ground Gaussian model (ablation) | ReconDreamer + depth loss · Waymo, lane shift 6m | NTL-IoU: 53.55 -> 55.34 | Improved |
| add separate ground Gaussian model (ablation) | ReconDreamer + depth loss · Waymo, lane shift 0m | PSNR: 35.13 -> 34.93 | Hurt |
| add NTDNet (ablation) | ReconDreamer + depth loss + ground model · Waymo, lane shift 0m | PSNR: 34.93 -> 36.29 | Improved |
| add NTDNet (ablation) | ReconDreamer + depth loss + ground model · Waymo, lane shift 6m | FID: 121.46 -> 111.92 | Improved |
| add NTDNet (ablation) | ReconDreamer + depth loss + ground model · Waymo, lane shift 0m | LPIPS: 0.1174 -> 0.1076 | Improved |
| full ReconDreamer++ versus ReconDreamer | ReconDreamer · Waymo, lane shift 0m | LPIPS: 0.1519 -> 0.1076 | Improved |
| full ReconDreamer++ versus ReconDreamer | ReconDreamer · Waymo, lane shift 3m | FID: 93.56 -> 72.02 | Improved |
| full ReconDreamer++ versus ReconDreamer | ReconDreamer · nuScenes, lane shift 3m | NTA-IoU: 0.325 -> 0.365 | Improved |
| full ReconDreamer++ versus ReconDreamer | ReconDreamer · EUVS test set (Level 1) | LPIPS: 0.3286 -> 0.3057 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- ReconDreamer — ReconDreamer: Crafting World Models for Driving Scene Reconstruction via Online Restoration (CVPR 2025) · better on 8 of 8 shared results
- Street Gaussians — Street Gaussians: Modeling Dynamic Urban Scenes with Gaussian Splatting (ECCV 2024) · better on 8 of 8 shared results
- FreeVS — FreeVS: Generative View Synthesis on Free Driving Trajectory (ICLR 2025) · better on 4 of 4 shared results
- NeuRAD — NeuRAD: Neural Rendering for Autonomous Driving (CVPR 2024) · better on 1 of 1 shared result
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
- ReconDreamer: Crafting World Models for Driving Scene Reconstruction via Online Restoration (CVPR 2025 · similarity 0.94)
- ReconDreamer-RL: Enhancing Reinforcement Learning via Diffusion-based Reconstruction (ECCV 2026 · similarity 0.90)
- DriveDreamer4D: World Models Are Effective Data Machines for 4D Driving Scene Representation (CVPR 2025 · similarity 0.86)
- FreeVS: Generative View Synthesis on Free Driving Trajectory (ICLR 2025 · similarity 0.86)
- ReSimAD: Zero-Shot 3D Domain Transfer for Autonomous Driving with Source Reconstruction and Target Simulation (ICLR 2024 · similarity 0.86)
- SceneCrafter: Controllable Multi-View Driving Scene Editing (CVPR 2025 · similarity 0.86)
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.
