Research map / Probabilistic and scientific modeling / Dynamical systems
Node-RF: Learning Generalized Continuous Space-Time Scene Dynamics with Neural ODE-based NeRFs
Hiran Sarkar, Liming Kuang, Yordanka Velikova, Benjamin Busam
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 pose and velocity auxiliary losses (ablation) | Node-RF · Oscillating Ball | IoU: 0.2730 -> 0.3253 | Improved |
| add pose and velocity auxiliary losses (ablation) | Node-RF · Oscillating Ball | PSNR: 28.661 -> 29.080 | Improved |
| add Lipschitz regularization (ablation) | Node-RF · Oscillating Ball | IoU: 0.3253 -> 0.3327 | No clear effect |
| add Lipschitz regularization (ablation) | Node-RF · Oscillating Ball | PSNR: 29.080 -> 29.091 | No clear effect |
| use latent size 512 instead of 256 (ablation) | Node-RF · Bouncing Balls (reconstruction split) | PSNR: 32.29 -> 33.70 | Improved |
| use latent size 512 instead of 1024 (ablation) | Node-RF · Bouncing Balls (reconstruction split) | PSNR: 32.74 -> 33.70 | Improved |
| use latent size 1024 instead of 256 (ablation) | Node-RF · Bouncing Balls (reconstruction split) | LPIPS: 0.0318 -> 0.0397 | Hurt |
| Node-RF instead of D-NeRF | dynamic NeRF · Pendulum (interpolation) | PSNR: 13.906 -> 17.057 | Improved |
| Node-RF instead of D-NeRF | dynamic NeRF · Pendulum (extrapolation) | PSNR: 13.295 -> 15.920 | Improved |
| Node-RF instead of D-NeRF(c) | conditioned dynamic NeRF · Oscillating Ball | IoU: 0.0008 -> 0.3327 | Improved |
| Node-RF instead of D-NeRF(c) | conditioned dynamic NeRF · Bifurcating Hill | IoU: 0.003 -> 0.485 | Improved |
| Node-RF instead of HexPlane | extrapolation · Bouncing Balls (4x extrapolation) | Motion Smoothness: 0.99617 -> 0.99648 | Mixed |
Closest papers by meaning
- BLiRF: Bandlimited Radiance Fields for Dynamic Scene Modeling (AAAI 2024 · similarity 0.89)
- Joint Optimization of Neural Radiance Fields and Continuous Camera Motion from a Monocular Video (CVPR 2025 · similarity 0.88)
- Ced-NeRF: A Compact and Efficient Method for Dynamic Neural Radiance Fields (AAAI 2024 · similarity 0.88)
- Gear-NeRF: Free-Viewpoint Rendering and Tracking with Motion-aware Spatio-Temporal Sampling (CVPR 2024 · similarity 0.87)
- Mask-Based Modeling for Neural Radiance Fields (ICLR 2024 · similarity 0.87)
- NeRF On-the-go: Exploiting Uncertainty for Distractor-free NeRFs in the Wild (CVPR 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.
