Research map / 3D vision / Gaussian splatting
HDGS: Hierarchical Dynamic Gaussian Splatting for Urban Driving Scenes
Fudong Ge, Jin Gao, Hanshi Wang, Yiwei Zhang, Ke Wang, Weiming Hu, Zhipeng Zhang
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 inter-anchor consistency enhancement (ablation) | HDGS · Waymo ablation | PSNR: 27.23 -> 28.04 | Improved |
| add global depth supervision (ablation) | HDGS · Waymo ablation | PSNR: 27.38 -> 28.04 | Improved |
| add local depth supervision (ablation) | HDGS · Waymo ablation | PSNR: 27.50 -> 28.04 | Improved |
| add gradient and density based growing (ablation) | HDGS · Waymo ablation | PSNR: 27.55 -> 28.04 | Improved |
| add unvisited-record pruning (ablation) | HDGS · Waymo ablation | PSNR: 27.98 -> 28.04; storage 303 -> 268 | Mixed |
| add global depth supervision to Gaussian DeSiRe-GS (ablation) | DeSiRe-GS · Waymo ablation | PSNR: 27.03 -> 27.05 | No clear effect |
| add feature modulation in ICE (ablation) | HDGS · Waymo ablation | PSNR: 27.39 -> 28.04 | Improved |
| add gated residual fusion in ICE (ablation) | HDGS · Waymo ablation | PSNR: 27.78 -> 28.04 | Improved |
| use 3 hierarchy levels instead of 1 (ablation) | HDGS · Waymo ablation | PSNR: 26.63 -> 28.04 | Improved |
| use 4 or 5 hierarchy levels instead of 3 (ablation) | HDGS · Waymo ablation | PSNR: 28.04 -> 25.72 (N=4), 25.40 (N=5) | Hurt |
| add depth supervision (ablation) | HDGS · KITTI depth | AbsRel: 0.242 -> 0.188 | Improved |
| replace Gaussian DeSiRe-GS with anchor-based HDGS | HDGS · Waymo SA NVS | storage (MB): 917 -> 284 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- DeSiRe-GS — DeSiRe-GS: 4D Street Gaussians for Static-Dynamic Decomposition and Surface Reconstruction for Urban Driving Scenes (CVPR 2025) · better on 7 of 7 shared results
- OmniRe — OmniRe: Omni Urban Scene Reconstruction (ICLR 2025) · better on 2 of 2 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
- Motion Decoupled 3D Gaussian Splatting for Dynamic Object Representation (AAAI 2025 · similarity 0.91)
- CityGaussianV2: Efficient and Geometrically Accurate Reconstruction for Large-Scale Scenes (ICLR 2025 · similarity 0.91)
- DeSiRe-GS: 4D Street Gaussians for Static-Dynamic Decomposition and Surface Reconstruction for Urban Driving Scenes (CVPR 2025 · similarity 0.91)
- HiCoM: Hierarchical Coherent Motion for Dynamic Streamable Scenes with 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.91)
- H3D-DGS: Exploring Heterogeneous 3D Motion Representation for Deformable 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- Efficient Gaussian Splatting for Monocular Dynamic Scene Rendering via Sparse Time-Variant Attribute Modeling (AAAI 2025 · similarity 0.91)
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.
