Research map / Graph learning / Knowledge and temporal graphs
ProSGNeRF: Progressive Dynamic Neural Scene Graph with Frequency Modulated Foundation Model in Urban Scenes
Tianchen Deng, Yanbo Wang, Yejia Liu, Chenpeng Su, Jingchuan Wang, Danwei Wang, Danwei Wang, Shao-Yuan Lo, Weidong Chen
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 progressive scene graph (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 26.34 -> 30.31 | Improved |
| add overlap between local scene graphs (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 28.79 -> 30.31 | Improved |
| add DINOv2 foundation encoder instead of MLP (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 28.75 -> 30.31 | Improved |
| add frequency modulation on top of DINOv2 (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 29.41 -> 30.31 | Improved |
| apply frequency modulation to MLP-only branch (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 28.75 -> 27.18 | Hurt |
| add object pose input (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 27.98 -> 30.31 | Improved |
| use separate encoders for shape and appearance (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 27.15 -> 30.31 | Improved |
| use separate networks for car and truck (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 27.81 -> 30.31 | Improved |
| add depth loss (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 29.25 -> 30.31 | Improved |
| add KL divergence loss (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 29.28 -> 30.31 | Improved |
| add plane sampling (ablation) | ProSGNeRF · KITTI | image reconstruction PSNR: 29.68 -> 30.31 | Improved |
| add ray-box sampling (ablation) | ProSGNeRF · KITTI | novel view synthesis PSNR (50%): 25.82 -> 26.55 | Improved |
Closest papers by meaning
- Multi-Level Neural Scene Graphs for Dynamic Urban Environments (CVPR 2024 · similarity 0.89)
- Dynamic 3D Gaussian Fields for Urban Areas (NeurIPS 2024 · similarity 0.89)
- EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision (ICLR 2024 · similarity 0.86)
- DORSal: Diffusion for Object-centric Representations of Scenes $\textit{et al.}$ (ICLR 2024 · similarity 0.86)
- PM-INR: Prior-Rich Multi-Modal Implicit Large-Scale Scene Neural Representation (AAAI 2024 · similarity 0.86)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 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.
