Research map / 3D vision / 3D generation
NeoVerse: Enhancing 4D World Model with in-the-wild Monocular Videos
Yuxue Yang, Lue Fan, Ziqi Shi, Junran Peng, Feng Wang, Zhaoxiang 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 |
|---|---|---|---|
| remove opacity regularization (add it) (ablation) | NeoVerse reconstruction · DyCheck | PSNR: 10.86 -> 11.56 | Improved |
| add bidirectional motion modeling (ablation) | NeoVerse reconstruction · DyCheck | PSNR: 11.27 -> 11.56 | Improved |
| add the generation model on top of reconstruction (ablation) | NeoVerse · DyCheck | PSNR: 11.56 -> 14.59 | Improved |
| reconstruct from sparse key frames (11 instead of all 81) (ablation) | NeoVerse generation · VBench novel view generation | imaging quality / inference time: 61.51 -> 59.75 imaging quality; 28 s -> 20 s | Mixed |
| increase key frames from 11 to full frames (ablation) | NeoVerse generation · VBench | Subject Consistency: 88.43 -> 89.42 | Improved |
| use scalable monocular pipeline with degradation simulation instead of TrajectoryCrafter | novel-view video generation · VBench novel view generation | Imaging Quality: 54.59 -> 61.51 | Improved |
| apply feed-forward pose-free 4DGS instead of AnySplat | static reconstruction · Scannet++ (32 views) | PSNR: 22.79 -> 25.34 | Improved |
| use NeoVerse instead of ReCamMaster | novel-view video generation · VBench | Temporal Flickering / Motion Smoothness: 96.56 -> 96.51 (full frames); 98.86 -> 98.67 | No clear effect |
Later papers that compared with this method
1 later accepted paper reports NeoVerse as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Taylor-Gaussians-Flow: Towards Non-uniform Motion for Novel View Synthesis from Monocular Video (ICML 2026 · better on 9 of 9 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MonST3R — MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion (ICLR 2025) · 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
- DynamicVerse: A Physically-Aware Multimodal Framework for 4D World Modeling (NeurIPS 2025 · similarity 0.88)
- GFlow: Recovering 4D World from Monocular Video (AAAI 2025 · similarity 0.88)
- CAT4D: Create Anything in 4D with Multi-View Video Diffusion Models (CVPR 2025 · similarity 0.87)
- World from Motion: Generative Dynamic Gaussian Reconstruction from Monocular Video (NeurIPS 2026 · similarity 0.87)
- SV4D: Dynamic 3D Content Generation with Multi-Frame and Multi-View Consistency (ICLR 2025 · similarity 0.86)
- Controlling Space and Time with Diffusion Models (ICLR 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.
