Research map / Video / Video generation
Walking in the Implicit: Interactive World Exploration via Neural Scene Representation
Zhiqi Li, Chengrui Dong, Zhenhua Du, Hangning Zhou, Cong Qiu, Hailong Qin, Mu Yang, Dongxu Wei, Peidong Liu
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 |
|---|---|---|---|
| use NIS latent instead of latent video frames (ablation) | NeuWorld Stage-1 prior · Re10K | FVD: 88.03 -> 86.20 | Improved |
| use NIS latent instead of latent video frames (ablation) | NeuWorld Stage-1 prior · Re10K | R_dist (deg): 4.20 -> 3.26 | Improved |
| use NIS latent instead of latent video frames (ablation) | NeuWorld Stage-1 prior · Re10K | T_dist: 0.141 -> 0.157 | Hurt |
| use NIS latent instead of latent video frames (ablation) | NeuWorld Stage-1 prior · Re10K | prior training time (h): 78.0 -> 17.2 | Improved |
| increase NIS tokens L from 512 to 3072 (ablation) | NIS-VAE · Re10K NVS | PSNR: 25.896 -> 28.352 | Improved |
| increase NIS channel width D from 32 to 256 (ablation) | NIS-VAE · Re10K NVS | PSNR: 26.245 -> 26.822 | No clear effect |
| add unified partial-NIS conditioning instead of cross-attention (ablation) | NeuWorld · Re10K forward short-term | R_dist: 0.095 -> 0.030 | Improved |
| add unified partial-NIS conditioning instead of cross-attention (ablation) | NeuWorld · Re10K forward long-term | T_dist: 0.203 -> 0.153 | Improved |
| add unified partial-NIS conditioning instead of cross-attention (ablation) | NeuWorld · Re10K forward long-term | LPIPS: 0.663 -> 0.671 | Hurt |
| add anti-drift condition augmentation (ablation) | NeuWorld · Re10K forward long-term | T_dist: 0.680 -> 0.153 | Improved |
| add anti-drift condition augmentation (ablation) | NeuWorld · Re10K forward long-term | R_dist: 0.495 -> 0.109 | Improved |
| add geometry-aware retrieval instead of recent-only (ablation) | NeuWorld · Re10K cycle return | R_dist: 0.940 -> 0.145 | Improved |
Closest papers by meaning
- Image2Sim: Scaling Embodied Navigation via Generative Neural Simulator (NeurIPS 2026 · similarity 0.88)
- Beyond Pixel Histories: World Models with Persistent 3D State (ICML 2026 · similarity 0.88)
- CameraCtrl II: Dynamic Scene Exploration via Camera-controlled Video Diffusion Models (ICCV 2025 · similarity 0.88)
- Novel View Synthesis from A Few Glimpses via Test-Time Natural Video Completion (NeurIPS 2025 · similarity 0.87)
- Learning 3D Persistent Embodied World Models (NeurIPS 2025 · similarity 0.86)
- Navigation World Models (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.
