Research map / 3D vision / Gaussian splatting
SteerX: Creating Any Camera-Free 3D and 4D Scenes with Geometric Steering
Byeongjun Park, Hyojun Go, Hyelin Nam, Byung-Hoon Kim, Hyungjin Chung, Changick Kim
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 |
|---|---|---|---|
| resample early (0.8T-0.6T) instead of later steps (ablation) | SteerX on Mochi · PenguinVideoBenchmark | Dyn-MEt3R: 0.910 -> 0.929 | Improved |
| resample early instead of linear (ablation) | SteerX on Mochi · PenguinVideoBenchmark | Dyn-MEt3R: 0.918 -> 0.929 | Improved |
| resample early instead of later (ablation) | SteerX on HunyuanVideo · PenguinVideoBenchmark | Dyn-MEt3R: 0.931 -> 0.964 | Improved |
| resample early instead of linear (ablation) | SteerX on HunyuanVideo · PenguinVideoBenchmark | Dyn-MEt3R: 0.943 -> 0.964 | Improved |
| resample later instead of early (ablation) | SteerX on Mochi · PenguinVideoBenchmark | Aesthetic: 0.500 -> 0.488 | Hurt |
| geometric steering instead of best-of-N | CogVideoX k=4 · VBench-I2V | Dynamic: 0.143 -> 0.170 | Improved |
| geometric steering instead of best-of-N | DimensionX k=4 · VBench-I2V Image-to-3D | GS-MEt3R: 0.745 -> 0.749 | Improved |
| geometric steering instead of best-of-N | SplatFlow k=4 · T3Bench | BRISQUE: 17.2 -> 13.1 | Improved |
| geometric steering instead of best-of-N | SplatFlow k=2 · T3Bench multi-view | BRISQUE: 36.82 -> 27.47 | Improved |
| increase particles from 4 to 8 (ablation) | SteerX on Mochi · PenguinVideoBenchmark | Dyn-MEt3R: 0.929 -> 0.945 | Improved |
| increase particles from 4 to 8 (ablation) | SteerX on HunyuanVideo · PenguinVideoBenchmark | Dyn-MEt3R: 0.964 -> 0.979 | Improved |
| increase frames from 25 to 49 (ablation) | SteerX on HunyuanVideo · PenguinVideoBenchmark | Dyn-MEt3R: 0.964 -> 0.978 | Improved |
Closest papers by meaning
- Free4D: Tuning-free 4D Scene Generation with Spatial-Temporal Consistency (ICCV 2025 · similarity 0.86)
- GenXD: Generating Any 3D and 4D Scenes (ICLR 2025 · similarity 0.85)
- Simulation-Ready Compositional 3D Scene Reconstruction from a Single Image (NeurIPS 2026 · similarity 0.84)
- DreamScene4D: Dynamic Multi-Object Scene Generation from Monocular Videos (NeurIPS 2024 · similarity 0.84)
- 3D-SceneDreamer: Text-Driven 3D-Consistent Scene Generation (CVPR 2024 · similarity 0.84)
- X-Scene: Large-Scale Driving Scene Generation with High Fidelity and Flexible Controllability (NeurIPS 2025 · similarity 0.84)
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.
